Sunteți pe pagina 1din 313

A e

g
ories of Readin
g
Figal Langage
in Russeau, Nietzsche, RiIe, ad Poust
Pau de Man
New Haven and London
Yale Universit Pess
1979
Pblished wth assistance from the Kngsley Tt A
siation Publication Fnd established b the Sroll and
K Siet of Yale College.
Copght 1979 by Yale Universit. All rghts reered.
This bok may not b reprouced, in whole or in part,
in any form (bond that copng prmitted in Sctions
107 and 108 of the U.S. Copyrght Law and except by
sion from the publishers.
Dsigned b Thos. Whitridge and set in Zapf Int
.
tional te. Printed in the United States of Are e by '
=
The Muray Pinting Company, Westford, Mas
,
setts:
"

Pblished in Great Brtain, Europe, Africa, and A ex-
cept Japan) by Yale Universit Press, Ltd., London. c
,
. "
tributed in Australia and New Zealand b Bk & Film
Srces, Artarmon, N.S.W., Australia; and in Japan by
Harr & Rw, Publishers, Tokyo Ofce.
Liba o ConCatagng in Pblication Data
De Man, Paul.
Alegories of rading.
Includes index.
1. French literature-Histor and criticism.
2. Russau, Jean Jacques, 1712-1778-Stle.
3
. Grman literatureHistor and criticism.
4. Figes of spech. 5. Alegor. I. Title.
PQ145.D5 80 79-6075
IS Q-30 02322 7
Qur on lit tp vi ou tp ducmnt on n 'ete rn.
Pascal
Contents
Peface i
Par I Rhetoric
1. Semiolog and Retoric 3
2. Tropes (Rile) 20
3. Rading (Proust) 57
4. Gnesis and Gnealog (Nietzsche) 79
5. Rhetoric of Tropes (Nietzsche) 103
6. Rhetoric of Persuasion (Nietzsche) 119
Par I Rousseau
7. Metaphor (Scon Dicourse) 135
8. Self (Plin) 16
9. Allegor Uuli) 188
10. Allegor of Reading (Pofesin d foi) 221
11. Promises (Scil Contat) 26
12. Ecuses (Conein) 278
Inde
303
Peface
ALLEGORIES OF READING STARTED OUT AS A HISTORICAL
study and ended up as a theor of reading. I began to read Russeau
seriously in preparation for a historical refection on Rmanticism
and found myself unable to progess beond local difculties of
interpretation. In trng to cope wth this, I had to shif from histori
cal defnition to the problema tics of reading. This shif, which is
tpical of my generation, is of more interest in its results than in its
causes. It could, in principle, lead to a rhetoric of reading reaching
beond the canonical principles of literar histor which still sere, in
this book, as the starting pint of their ow displacement. The prn
ciples underlyng the thematic diversit of Russeau, the chronolog
of RIe and Nietzsche, the rhetoric of Proust, are not lef intact by
the reading, but this critical result remains dependent on the initial
position of these ver principles. Wether a further step, which
would leave this hermeneutic model behind, can b taken should not
a prr or naively be taken for ganted.
In Part II, on Russeau, I have attempted the elaboration and
the undoing of a system of tropological transformations in the form
of a sustained argment. Part I establishes a similar pattern in a
more fragmented way by moving between several authors rather
than staying wthin a single corus. The choice of Proust and ofRIe
as examples is partly due to chance, but since the ostensible pathos of
their tone and depth of their statement make them particularly resis
tant to a reading that is no longer entirely thematic, one could arge
that if thir work yields to such a rhetorcal scheme, the same would
necessarily be tre for witers whose rhetorical strategies are less
hidden behind the seductiv powers of identifcation.
Wat emerges is a process of reading in which rhetoric is a
dis
rptive intertning of trope and persuasion or-which is not
qu
ite the sme thing-of cognitive and performative langage. The
i
mplications of this conclusion ar not easy to unfold, nor can the be
stated in summar fashion, separated from the intricacies of spcifc
rea
dings. Nevertheless, opponents of such an apprach have been
more eager to attack what the assume to be its ideological motives
Ix
PREFACE
rather tha the technicalites of its proue. This is parculaly tre
wth regrd to the ten "deconstrction," which has rapidy bcome
a label as well as a target. Most of this bok was witten before
"deonstruction" bcame a bne of contention, and the term is used
here in a tehnical rather than a polemical sensewhich dos not
imply that it therefore becomes neutral or ideologically innoent. But
I sw no reason to delete it. No other word states so economically the
imp ibilit to evaluate positively or negtively the inescapable
evaluation it implies. Something is lost when the same proess is
described by a purely negtive term, as when Nietzsche sp of the
destruction (Zmmrng) of conceptual constructs or Pascl of
the demolition (dmliton) of a conviction that is itself already a
destruction. I consciously came across "deconstruction" for the frst
time in the witing of Jacques Derida, which means that it is
associated with a power of invntive rigor to which I lay no claim but
which I certainly do not wsh to erase. Deconstruction, as was easily
predictable, has been much misrepresented, dismissed as a harmless
academic game or denounced as a terrorist weapon, and I have all
the fewer illusions about the possibilit of countering these aberra
tions since such an exectation would go against the drif of my ow
readings.
Allr q Rang was a long time in the writing, and the list
of institutions to which I am indebted is even longer. I began to write
on Russau and Nietzsche with the assistance of a Gugenheim
Fellowship in 1969 and wte the main part of the book during a
year's leave from Yale University in 1972-73, wth the assistance of
a Yale Snior Faculty Fellowship supplemented by a gant fom the
Merrit Foundation and a grant-in-aid from the Aerican Council for
Larned Soieties. Final verifcations were completed in 1978 with
the help of a travel grant from the Griswold Fund at Yale. I wsh to
thank the numerous colleages whose support helped me in securing
this aid. A for my intellectual indebtedness, I feel indeed unable to
enumerate what is beond numbr and to disentangle, in so many
cases, the part of influence fom the part of friendship.
Parts of this book have appared in print bfore. The setion on
Proust was orignally a contribution to a Fetchrif for Gorges
Poulet entitled Mouvement preis (Paris:Jose Corti, 1972) and the
Rlke chapter was written as an introduction to the French edition of
Rlke's poms (Paris: Editions du Suil, 1972). Other chapters ap
peared entirely or in part in Critcal Inquir, Dirit, The Gr
PREFACE
R, Glyph, Stui in Rmantm, and Yal Fh Stui. Per
mission to reprint is grateflly ackowledged. I have myself trans
lated the to sections orignally witten in French.
I wish to than Ellen Graham and Sheila Huddleston of the Yae
University Press for particularly efcient and speedy copyediting,
ce
rtain to cleanse the fnal tet of all mistakes but my ow.
All translations of French and Gr quotations are my ow
unless otherse indicated.
P. d. M.
Ne Haven, April 1979
Pa I
Retorc
1
Semiolog
and Retorc
TO JUDGE FROM VARIOUS RECENT PUBLICATIONS, THE
spirit of the times is not blowing in the direction of formalist and
intrinsic criticism. We may no longer be heang too much about
relevance but we keep hearing a great deal about reference, about
the nonverbal "outside" to which langage refers, by which it is
conditioned and upon which it acts. The stress falls not so much on
the fctional status of literature-a propert now perhaps somewhat
too easily taken for ganted-but on the interplay beteen these
fctions and categories that are said to partake of realit, such as the
self, man, societ, "the artist, his cultue and the human commu
nit," as one critic puts it. Hence the emphasis on hybrid texts con
sidered to be partly literar and partly referential, on popular fctions
deliberately aimed towards social and psychological gratifcation; on
literar autobiography as a key to the understanding of the self, and
so on. We speak as if, with the problems of literar form resolved
once and forever, and with the techniques of structural analysis
refned to near-perfection, we could now move ''beyond formalism"
towards the questions that really interest us and reap, at last, the
fruits of the ascetic concentration on techniques that prpared us for
this decisive step. With the internal law and order of literature well
policed, we can now confdently devote ourselves to the foreign af
fairs, the external politics of literature. Not only do we feel able to do
so, but we owe it to ourselves to tae this step: our moral conscience
would not allow us to do otherwise. Behind the assurance that valid
interpretation is possible, behind the recent interest in wrting and
reading as potentially efective public speech acts, stands a highly
respectable moral imperative that strives to reconcile the interal,
formal, private structues of literar language with their external,
referential, and public efects.
I want, for the moment, to consider brefy this tendenc in
itself, as an undeniable and recurent historical fact, without regard
3
RHETORIC
for its truth or falseness or for its value as desirable or pernicious. It
is a fact that t
his sort of thing
happens, again and again, in literar
studies. On the one hand, literature cannot merely be received as a
defnite unit of referential meaning that can be decoded wthout
leaving a residue. The code is unusually conspicuous, complex, and
enigatic; it attracts an inordinate amount of attention to itself, and
this attention has to acquire the rigor of a method. The structural
moment of concentration on the code for its ow sake cannot be
avoided
, and literature necessarily breeds its ow formalism. Tech
nical innovations in the methoical study of literature ony occur
when this knd of attention predominates. It can legitimately be said,
for example,
that, from a technical point of vew, ver little has
happen
ed in
American criticism since the innovative work of New
Criticism. There certainly have been numerous excellent book of
criticism since, but in none of them have the techniques of descrip
tion and interpretation evolved beyond the techniques of close read
ing established in the thirties and the forties. Formalism, it seems, is
an all-absorbing and tyrannical muse; the hope that one can be at
the same time technically original and discursively eloquent is not
borne out by the histor of literar criticism.
On
the other hand-and this is the real myster-no literar
formalism, no matter how accurate and enriching in its analytic
powers, is evr allowed to come into being wthout seeming reduc
tive. When form is considered to be the external trappings of literar
meaning or content, it seems supercial and expendable. The de
velopment of intrinsic, formalist criticism in the twentieth centur
has changed this model: form is now a solipsistic categor of self
reflection, and the referential meaning is said to be extrinsic. The
polarities of inside and outside hav been reversed, but they are still
the same polarities that are at play: internal meaning has become
outside reference, and the outer form has become the intrinsic struc
ture. A new vrsion of reductivness at once follows this reversal:
formalism nowadays is mostly described in an imager of impris
onment and claustrophobia: the "prison house of language," "the
impasse of formalist criticism," etc. Like the grandmother in Proust's
novel ceaselessly driving the young Marcel out into the grden, away
from the unhealthy inwardness of his closeted reading, critcs cr out
for the fresh air of referential meaning. Thus, wth the structure of
the code so opaque, but the meaning so anxous to blot out the
ob
stacle of form, no wonder that the reconciliation of form and
SEMIOLOGY AND RHETORIC
I
meaning woud b so attractive. The attraction of reconciliation is
the electve breeding-groud of false models and metaphors; it ac
couts for the metaphorical moel of literature as a knd of box that
separates an inside from an outside, and the reader or critic as the
person who opens the lid in order to release in the open what was
secreted but inaccessible iside. It matters little whether we call the
inside of the box the content or the form, the outside the meaning or
the appearance. The recurent debate opposing intrinsic to extrinsic
criticism stands under the aegis of an inside/outside metaphor that is
never being seriously questioned.
Metaphors are much more tenacious than facts, and I certainly
don't expect to disloge this age-old model in one short tr. I merely
wsh to speculate on a diferent set of terms, perhaps less simple in
their diferential relationships than the strictly polar, binar opposi
tion between inside and outside and therefore less likely to enter into
the easy play of chiasmic reversals. I deriv these terms (which are as
old as the hills) pragmatically from the obseration of developments
and debates in recent critical methodolog.
One of the most controversial among these developments coin
cides wth a new approach to poetics or, as it is called in Grmany,
poetolog, as a branch of general semiotics. In France, a semiolog of
literature comes about as the outcome of the long-deferred but all
the more exlosive encounter of the nimble French literar mind
wth the categor of form. Semiolog, as opposed to semantics, is the
science or study of signs as signifers; it does not ask what words
mean but how the mean. Unlike Aerican New Criticism, which
derived the internalization of form from the practice of highly self
conscious modern writers, French semiolog turned to linguistics for
its model and adopted Saussure and Jakobson rather than Valer or
Proust for its masters. By an awareness of the arbitrariness of the
sign (Saussue) and of literature as an autotelic statement "focused
on the way it is expressed" Uakobson) the entire question of meaning
can be bracketed, thus freeing the critical discourse from the de
bilitatng buden of paraphrase. The demysti:ng power of semiol
og, wthin the context of French historical and thematic criticism,
has been considerable. It demonstrate that the perception of the
literar dimensions of language is largely obscured if one submits
uncritically to the authority of reference. It also revealed how tena
ciously this authority continues to assert itself in a variety of dis
guises, ranging from the crudest ideolog to the most refned forms

RHETORIC
of aesthetc ad ethca judgent. It especially explodes the myh of
smtc cores
p
ndnce bteen sign and referent, the wshfl
hop of hvg it bth ways, of bing, to paraphrase Mar in the
Gra Ideolog, a formalist critic in the morning and a communal
moralit in the aferoon, of serg both the technique of form and
the substance of meaning. The results, in the practice of French
criticism, hav been as fruitfl as the are ireversible. Perhaps for
the fst tme since the late eighteenth centur, French crtics can
come at least somewhat closer to the kind of linguistic awareness
that never ceased to b oprative in its pets and novelists and that
forced all of them, including Sinte Beuve, to wrte their main work
"contre Sainte Beuve." The distance was never so considerable in
England and the United States, which does not mean, however, that
we may be able, in this countr, to dispense altogether wth some
preventative semiologcal hygene.
One of the most strng characteristics of literar semiolog as
it is practiced today, i Fance and elsewhere, is the use of grammat
ical (especially syntactical) structures conjointly wth rhetorcal
structures, wthout apparent awareness of a possible discrepanc
beteen them. In their literar analyses, Barthes, Gnette, Toorov,
Greimas, and their disciples all simpli: and regress from Jakbson
in lettng gammar and rhetoric function in perfect continuit, and
in passing from gammatical to rhetorcal structures wthout
difcult or interption. Indeed, as the study of gammatical struc
tures is refned in contemporar theories of generative, transforma
tional, and distributive grammar, the study of tropes and of fgures
(which is how the term rhtori is used here, and not in the derived
sense of comment or of elouence or prsuasion) becomes a mere
extension of grammatical moels, a particular subset of syntactical
relations. In the recent Ditionnaire enclpeu ds sin du
lngage, Ducrot and Todorov write that rhetoric has always been
satisfed wth a paradigmatic vew over words (words substituting
for each other), wthout questioning their syntagmatic relationship
(the contiguit of words to each other). There ought to be another
prspective, complementar to the frst, in which metaphor, for
example, would not be defned as a substitution but as a particular
t of combination. Rsearch inspired by linguistics or, more nar
rowly, b syntactical studies, has begun to reveal this possibilit
but it remains to b explored. Toorov, who calls one of his bok a
Grmmr of th Deamrn, rightly thinks of his own work and that
SEMIOLOGY AND RHETORIC 7
of his associates as frst explorations in the elaboration of a system
atic grammar of literar modes, genres, and also of literar fgures.
Perhaps the most perceptive work to come out of this school, G
nette's studies of fgural modes, can be show to be assimilations of
rhetorical transformations or combinations to sytactical, grammat
ical patterns. Thus a recent study, now prnted in Figre I and
entitled Metphr an Metnymy in Put, shows the combined
presence, in a wde and astute selection of passages , of paradigmatic ,
metaphorical fgures wth stagmatic, metonyic structures. The
combination of both is treated descriptively and nondialectically
without considering the possibilit of logical tensions.
One can ask whether this reduction of fgure to grammar is
legtimate. The existence of grammatical structures, wthin and be
yond the unit of the sentence, in literar texts is undeniable, and their
description and classifcation are indispnsable. The question re
mains if and how fgres of rhetorc can be included in such a
taxonomy. Tis question is at the core of the debate going on, in a
wde variet of apparently unrelated forms, in contemporar poetics.
But the historical picture of contemporar criticism is too confused
to make the mapping out of such a topogaphy a useful exercise. Not
only are these questions mixed in and mixed up wthin particular
groups or local trends, but the are ofen co-present, wthout appar
ent contradiction, wthin the work of a singe author.
Neither is the theor of the question suitable for quick exposi
tor treatment. To distingish the epistemolog of grammar from
the epistemolog of rhetoric is a redoubtable task. On an entirely
naive level, we tend to conceive of gammatical sstems as tending
towards universalit and as simply generative, i.e., as capable of
derivng an innit of versions from a singe moel (that may govern
transformations as well as derivations) wthout the interention of
another moel that would upset the frst. We therefore think of the
relationship between grammar and logic, the passge from gram
mar to propositions, as being relatively unproblematic: no tre prop
ositions are conceivable in the absence of gammatical consistenc or
of controlled deviation from a system of consistenc no matter how
complex. Grammar and logc stand to each other in a dyadic rela
tionship of unsubverted support. In a logc of acts rather than of
statements, as in Austin's theor of speech acts, that has had such a
strong influence on recent American work in literar semiolog, it is
also possible to move btween speech acts and grammar wthout
8
RHETORIC
difculty. The
p
r
on
ance of w
hat is called illocutionar acts such
as orderig,
q
u
estioning, denying, assuming, etc., wthin the lan
guage is
con
g
ent
w
ith the grammatical structures of syntax in the
corres
p
nding im
p
rative
, interrogative, negtive, optative sen
tences.
"
The
rles for illoc
utionar acts," wites Rchard Ohman in a
recent
p
per, "determine whether per
ormance of a given act is
well-executed, in just the same way as gammtial rules determine
whether the
product of a locutiona act-a sentence-is well
formed. . . . But whereas the rles o gammar concer the rla
tionships among sound, sytax, and meaning, the rules of illocu
tionar acts concern relationships among people."l And since rhet
oric is then conceived exclusively as persuasion, as actual action upon
others (and not as an intralinguistic fgure or trope), the continuit
between the illocutiona realm of grammar and the perlocutionar
realm of rhetoric is self-evdent. It becomes the basis for a new
rhetoric that, exactly as is the case for Todorov and Gnette, would
also be a new gammar.
Without engaging the substance of the question, it can be
pointed out, wthout having to go beyond rcent and Aercan
examples, and wthout calling upon the strength of an age-old tradi
tion, that the continuity here assumed btween grammar and
rhetoric is not borne out by theoretical and philosophical specula
tion. Kenneth Burke mentions dfctn (which he compares strc
turally to Freudian displacement), defned as "any slight bias or even
unintended error ," as the rhetorical basis of language, and defection
is then conceived as a dialectical subversion of the consistent link
between sign and meaning that operates wthin grammatical pat
terns; hence Burke's well-kown insistence on the distinction be
tween grammar and rhetorc. Charles Sanders Peirce, who, wth
Nietzche and Sussure, laid the philosophical foundaton for moder
semiolog, stressed the distinction between grammar and rhetoric in
his celebrated and so sugestively unfathomable defnition of the
sign. He insists, as is well know, on the necessar presence of a third
element, called the interretant, within any relationship that the sign
entertains with its object. The sign is to be interpreted if we are to
understand the idea it is to conve, and this is so because the sign is
not the thing but a meaning derved from the thing by a process here
1. "Spech, Literature, and the Space in Between," Ne Litrar Hitor 4
(Autumn 197): 50.
SEMIOLOGY AND RHETORIC 9
called representation that is not simply gnerativ, Le., dependent on
a univocal origin. The interpretation of the sign is not, for Peirce, a
meaning but another sign; it is a reading, not a decodage, and this
reading has, in its tun, to be interpreted into another sign, and so on
a infnitum. Peirce calls this process b means of which "one sign
gives birth to another" pue rhetoric, as distingshed from pue
grammar, which postulates the possibility of unproblematic, dyadic
meaning, and pure logic, which postulates the possibility of the
universal truth of meanings. Only if the sign engendered meaning in
the same way that the object engenders the sign, that is, by repre
sent
ation, would there be no need to distinguish between gammar
and rhetoric.
These remarks should indicate at least the exstence and the
difculty of the question, a difculty which puts its concise theoreti
cal exposition beond my powers. I must retreat therefore into a
pragmatic discouse and tr to illustrate the tension between gram
mar and rhetoric in a few specifc textual examples. Let me begin b
considerng what is perhaps the most commonly kown instance of
an apparent symbiosis between a gammatical and a rhetorical
strctue, the so-called rhetorical
question, in which the fgure is
conveyed directly b means of a syntactical device. I take the frst
example from the sub-literatue of the mass media: asked by his wfe
whether he wants to have his bowling shoes laced over or laced
under, Archie Bunker answers wth a question: "What's the difer
ence?" Being a reader of sublime simplicity, his wfe replies by pa
tiently explaining the diference between lacing over and lacing
under, whatever this may be, but provokes only ire. "Wat's the
diference" did not ask for diference but means instead "I don't give
a damn what the diference is." The same gammatical patter
engenders two meaning that are mutually exclusive: the literal
meaning asks for the concept (diference) whose exstence is denied
b the fgrative meaning. As long as we are talking about bowling
shoes, the consequences are relatively trvial; Archie Bunker, who is a
geat believer in the authorty of origns (as long, of course, as the
are the rigt origins) muddles along in a world where literal and
fgurative meaning get in each other's way, though not wthout
discomforts. But suppose that it is a d-buner rather than a
"
B
unker," and a de-bunker of the arche (or origin), an archie De
buner such as Nietzsche or Jacques Derrida for instance, who ask
the question "What is the Diference"-and we cannot even tell from
10
RHETORIC
his
gammar
whether he "really" wats to k
n
ow
"what" diference is
or
is
just telling us that we shouldn't even tr to fnd out. Confronted
wth the question of the diference
b
teen
g
a
mmar and rhetoric,
g
ammar alows us to ask the question, but the sentence by means of
which w ask it may deny
the ve
r possibilit of askng. For what is
the use of askng, I ask, when we cannot even authoritatively decide
whether a question ask or doesn't ask?
The point is as follows. A perfectly clear syntactical paradigm
(the question) engenders a sentence that has at least to meaning,
of which the one asserts and the other denies
i
ts ow illocutionar
mode. It is not so that there are simply to meanings, one literal and
.the other fgal, and that we have to decide which one of these
meanings is the rght one in this particuar situation. The confusion
can only be cleared up by the interention of an extra-textual inten
tion, such as Archie Buer putting his wife straight; but the ver
anger he displays is indicative of more than impatience; it reveals his
despair when confronted with a structure of linguistic meaning that
he cannot control and that holds the discouraging prospct of a
infnit of similar future confsions, all of them potentially cata
strophic in their consequences. Nor is this interntion really a part
of the mini-text constituted b the fge which holds ou attention
only as long as it remains suspended and uesolved. I follow the
usage of common speech in calling this semiological enigma "rhetor
ical." The grammatical model of the question becomes rhetorical not
when we have, on the one hand, a literal meaning and on the other
hand a fgural meaning, but when it is impossible to decide by
gammatical or other linguistic devices which of the two meanings
(that can be entirely incompatible) prevails. Retoric radicaly sus
pends logic and opens up vertignous possibilities of referential aber
ration. And although it would perhaps be somewhat more remote
from common usage, I would not hesitate to equate the rhetorical,
fgural potentialit of language wth literature itself I could point to
a geat number of antecedents to this equation of literature with
fgure; the most recent reference would be to Monroe Beardsley's
insistence in his contribution to the Eays to honor William Wim
stt, that literar langage is characterzed by being "distinctly above
the norm in ratio of implicit [or, I would say rhetorica] to explicit
meaning."2
2. "The Concept of Literature," in Litar Thr an Str: Es ays in
Honr of Wilim K Wimt, ed. Frank Brady,John Palmer, and Martin Pce (New
Haven, 197), p.
37.
SEMIOLOGY AND RHETORIC 11
Let me pursue the matter of the rhetorical question through one
more example. Yeats's poem "Among Shool Children" ends with the
famous line: "How can we kow the dancer from the dance? A
though there are some revealing inconsistencies within the commen
taries, the,line is usually interpreted as stating, with the increased
emphasis of a rhetorical devce, the potential unit beteen form
and experience, between creator and creation. It could be said that it
denies the discrepanc between the sign and the referent from which
we started out. Many elements in the imager and the dramatic
development of the poem strengthen this traditional reading; with
out having to look any further than the immediately preceding lines,
one fnds powerful and consecrated images of the continuit from
part to whole that makes synecdoche into the most seductive of
metaphors: the organic beaut of the tree, stated in the parallel
syntax of a similar rhetorical question, or the convergence, in the
dance, of erotic desir wth musical form:
o chestnut-tree, great-rooted blossomer,
Ae you the leaf, the blossom or the ble?
o by swayed to music, 0 brightening glance,
How can we kow the dancer from the dance?
A more extended reading, always assuming that the fnal line is to be
read as a rhetorical question, reveals that the thematic and rhetorical
grammar of the poem yields a consistent reading that etends from
the frst line to the last and that can account for all the details in the
text. It is equally possible, however, to read the last line literally
rather than fguratively, as askng with some ugenc the question
we asked earlier within the context of contemporar criticism: nt
that sign and referent are so exquisitely ftted to each other that all
diference between them is at times blotted out but, rather, since the
two essentially diferent elements, sign and meaning, are so intri
cately intertwined in the imagined "presence" that the poem ad
dresses, how can we pssibly mae the distinctions that would shel
ter us from the error of identitng what cannot be identifed? The
clumsiness of the paraphrase reveals that it is not necessrily the
literal reading which is simpler than the fgurative one, as was the
case in our frst example; here, the fgral reading, which assumes
the question to be rhetorical, is perhaps naive, whereas the literal
reading leads to geater complication of theme and statement. For it
turns out that the entire scheme set up by the frst reading can be
12 RHETORIC
undermi
e
d, or deconstrcted, in the terms of the second, in which
the fnal line is read literally as meaning that, since the dancer and
the dance are not the same, it might be useful, perhaps even desper
ately necessar-for the question can be given a ring of urgenc,
"Please tell me, how can I know the dan
cer from the dance"-to tell
them apart. But this wll replace the reading of each symbolic detail
by a divergent interretation. The oneness of trnk, leaf, and blos
som, for example, that would have appealed to Gthe, would fnd
itself replaced by the much less reassuring Tree of Life from the
Mabinogion that appeas in the poem "Vacillation," in which the
fer blossom and the earthly leaf are held together, as well as apart,
by the crcifed and castrated Gd Attis, of whose body it can hardly
be said that it is "not bruised to pleasure soul." This hint should
sufce to suggest that two entirely coherent but entirely incompati
ble readings can be made to hinge on one line, whose grammatical
strcture is devoid of ambiguity, but whose rhetorical mode turns
the mood as well as the mode of the entire poem upside down.
Neither can we sy, as was already the case in the frst example, that
the poem simply has two meanings that exst side b side. The two
reading have to engage each other in direct confrontation, for the
one reading is precisely the error denounced by the other and has to
be undone by it. Nor can we in any way make a valid decision as to
which of the readings can be given prority over the other; none can
exst in the other's absence. There can be no dance without a dancer,
no sign without a referent. On the other hand, the authority of the
meaning engendered by the grammatical strcture is flly obscued
by the duplicit of a fgure that cries out for the diferentiation that it
conceals.
Yeats's poem is not explicitly "about" rhetorical questions but
about images or metaphors, and about the possibilit of convergence
between experiences of consciousness such as memor or
emotions-what the poem calls passion, piety, and afection-and
entities accessible to the snses such as bodies, persons, or icons. We
retun to the inside/outside model from which we started out and
which the poem puts into question by means of a syntactical device
(the question) made to operate on a gammatical as well as on a
rhetorical level. The couple grammar/rhetoric, certainly not a binar
opposition since they in no way exclude each other, disrpts and
confuss the neat antithesis of the inside/outside pattern. We can
transfer this scheme to the act of reading and interretation. By
reading we get, as we say, inid a text that was frst something alien
SEMI OLOGY AND RHETORI C 13
to us and which we now mae our own b an act of understanding.
But this understanding becomes at once the representation of an
extra-textual meaning; in Austin's terms, the illocutionar speech act
becomes a prlocutionar actual act-in Frege's terms, Bdutng
becomes Sinn. Our recurent question is whether this transformation
is semantically controlled along grammatical or along rhetorical
lines. Does the metaphor of reading really unite outer meaning wth
inner understanding, action with refection, into one single totalit?
The assertion is powerfully and sugestively made in a passage from
Poust that describes the experience of reading as such a union. It
descrbes the young Marcel, near the begnning of Combray, hiding
in the closed space of his room in order to read. The example difers
from the earlier ones in that we are not dealing with a grammatical
strcture that also functions rhetorically but have instead the repre
sentation, the dramatization, in terms of the experience of a subject,
of a rhetorical structujust as, in many other passages, Proust
dramatizes tropes b means of landscapes or descriptions of objects.
The fge here dramatized is that of metaphor, an inside/outside
correspondence as represented by the act of reading. The reading
scene is the culmination of a series of actions takng place in enclosed
spaces and leading up to the "dark colness" of Marcel's room.
I had stretched out on my bed, wth a book, in my room which
sheltered, tremblingly, its transparent and fragle coolness from
the afernoon sun, behind the almost closed blinds through
which a glimmer of daylight had nevertheless managed to push
its yellow wngs, remaining motionless between the wood and
th gass, in a corner, poised like a buttery. It was hardy light
enough to read, and the sensation of the light's splendor was
given me only b the noise of Cmus . . . hammering dusty
crates; resounding in the sonorous atmosphere that is peculiar
to hot weather, the seemed to spark of scarlet stars; and also
b the flies executing their little concert, the chamber music of
summer: evocative not in the manner of a human tune that,
heard perchance duing the summer, aferwards reminds you
of it but connected to summer by a more necessar link: born
from beautiful days, resurrecting ony when the return, con
taining some of their essence, it dos not only awaken their
image in ou memor; it guarantees their return, their actual,
persistent, unmediated presence.
The dark coolness of my room related to the full sunlight of
14
RHETORI C
the str
eet
as the shadow relates to the ray of light , that is to say
it was
just as luminous ad it gave my imagination the total
spectacle of the sum er, whereas my senses, if! had been on a
wal, could ony have enjoyed it by fragments; it matched my
repose which (tha to the adventues told by my book and
stirring my tranquilit) supported, like the quiet of a motionless
hand i the mdde of a rning brook the shock and the motion
of a torrent of activt. [Swann
'
s Way. Paris: PlEiade, 1954, p.
83.]
For our present purpose, the most striking aspect of this passge
is the juxtaposition of fgual and metafgral langage. It contains
seductve metaphors that bring into play a variet of iresistible
objects: chamber music, butterfies, stars, book, running brook,
etc., and it inscribes these objects within dazzling fre- and water
work of fgration. But the passage also comments normatively on
the best way to achiee such efects; in this sense, it is metafgural: it
writes fgratively abut fgures. It contrasts two ways of evokng the
natural experience of summer and unambiguously states its prefer
ence for one of these ways over the other: the "necessar lin" that
unites the buzzing of the fies to the summer makes it a much more
efective symbol than the tune heard "perchance" duing the sum
mer. The preference is expressed by means of a distinction that
corresponds to the diference beteen metaphor and metonymy,
necessit and chance being a legitimate way to distinguish between
analog and continguit. The inference of identit and totalit that is
constitutive of metaphor is lacking in the purely relational meto
nymic contact: an element of truth is involved in takng Achilles
for a lion but none in takng Mr. Ford for a motor car. The passage is
aout the aesthetic superiorit of metaphor over metonymy, but this
aesthetic claim is made by means of categories that are the ontologi
cal ground of the metaphy ;ical system that allows for the aesthetic to
come into being as a ca'cegor. The metaphor for summer (in this
case, the synesthesia set of by the "chamber music" of the fies)
guarantees a presence which, far from bing contingent, is said to be
essential, permanently recurrent and unmediated by lingistic repre
sentations or fgurations. Finally, in the second part of the passage,
the metaphor of presence not only appears as the ground of cogni
tion but as the perormance of an action, thus promising the recon
ciliation of the most disruptive of contradictions. By then, the in-
SEMI OLOGY AND RHETORI C 15
vestment in the power of metaphor is such that it may seem sac
rilegious to put it in question.
Yet, it takes little perspicacit to show that the text does not
practice what it preaches. A rhetorical reading of the passage reveals
that the fgural praxis and the metafgural theor do not converge
and that the assertion of the master of metaphor over metonyy
owes its persuasive power to the use of metonymic structures. I have
carried out such an analysis in a somewhat more extended context
(pp. 59-67, below); at this point, we are more concerned wth the
results than wth the procedure. For the metaphysical categories of
presence, essence, action, truth, and beaut do not remain unaf
fected by such a reading. This would become clear from an inclu
sive reading of Proust's novel or would become even more explicit
in a language-conscious philosopher such as Nietzsche who, as a
philosopher, has to be concerned wth the epistemological conse
quences of the knd of rhetorical seductions exemplifed by the Proust
passage. It can be show that the systematic critique of the main
categories of metaphysics undertaken by Nietzsche in his late work,
the critique of the concepts of causality, of the subject, of identit, of
referential and revealed truth, etc., occurs along the same pattern of
deconstruction that was operative in Proust's text; and it can also be
show that this pattern exactly corresponds to Nietzsche's descrip
tion, in texts that precede Th Will t Powr by more than ffeen
years, of the structure of the main rhetorical tropes. The key to this
critique of metaphysics, which is itself a recurrent gesture through
out the histor of thought, is the rhetorical model of the trope or, if
one prefers to call it that, literature. It turns out that in these
innocent-lookng didactic exercises we are in fact playng for ver
sizeable stakes.
It is therefore all the more necessr to kow what is linguisti
cally involved in a rhetorically conscious reading of the type here
undertaken on a brief fragment from a novel and extended by
Nietzsche to the entire text of post-Hellenic thought. Our frst exam
ples dealing wth the rhetorical questions were rhetorizations of
grammar, fgures generated by syntactical paradigms, whereas the
Proust example could be better described as a grammatization of
rhetoric. By passing from a paradigmatic structure based on sub
stitution, such as metaphor, to a syntagmatic strcture based on
contingent association such as metonymy, the mechanical, repetitive
aspect of grammatical forms is show to be operative in a passage
16
RHETORI C
that seemed at frst sight to celebrate the self-wlled and autonomous
inventiveness of a subject . Figures are assumed to be inventions, the
proucts of a highly particularized individual talent , whereas no one
can claim credit for the programmed pattern of grammar. Yet , our
reading of the Proust passage shows that precisely when the highest
claims are being made for the unifng power of metaphor, these
ver images rely in fact on the deceptive use of semi-automatic
grammatical patterns. The deconstruction of metaphor and of all
rhetorical patterns such as mimesis, paranomasis, or prsonifcation
that use resemblance as a way to disguise diferences, takes us back
to the impersonal precision of gammar and of a semiolog derived
from grammatical patterns. Such a reading puts into question a
whole series of concepts that underlie the value j udgments of our
crtical discourse: the metaphors of primac, of genetic histor, and,
most notably, of the autonomous power to will of the self.
There seems to b a diference, then, between what I caled the
rhetorization of grammar (as in the rhetorical question) and the
grammatization of rhetoric, as in the readings of the type sketched
out in the passage from Proust. The former end up in indetermina
tion, in a suspended uncertaint that was unable to choose between
two modes of reading, whereas the latter seems to reach a truth,
albeit by the negative road of exposing an error, a false pretense.
Afer the rhetorical reading of the Proust passge, we can no longer
believe the assertion made in this passage about the intrinsic,
metaphysical superiorit of metaphor over metonymy. We seem to
end up in a mood of negative assurance that is highly productive of
crtical discourse. The further text of Proust's novel , for example,
responds perfectly to an extended application of this pattern: not
only can similar gestures be repeated througout the novel, at all the
crucial articulations or all passages where large aesthetic and
metaphysical claims are being made-the scenes of involuntar
memor, the workhop of Elstir, the septette of Vinteuil , the con
vergence of author and narrator at the end of the novel-but a vast
thematic and semiotic network is revealed that structures the entire
narrative and that remained invisible to a reader caught in naIve
metaphorical mystifcation. The whole of literature would respond
in similar fashion, although the techniques and the patterns would
have to var considerably, of course, from author to author. But there
is absolutely no reason why analyses of the knd here suggested for
Proust would not be applicable, wth proper moifcations of tech-
SEMI OLOGY AND RHETORI C 17
nique, to Milton or to Dante or to Holderlin. This wll in fact be the
task of literar criticism in the coming yeas.
It would seem that we are saying that criticism is the decon
struction of literature, the reduction to the rigors of gamma of
rhetorical mystifcations. Ad if we hold up Nietzsche as the
philosopher of such a critical deconstruction, then the literar critic
would become the philosopher's ally in his strugle wth the pots.
Criticism and literature would separate around the epistemological
axis that distinguishes gammar from rhetoric. It is easy enoug to
see that this apprent gorifcation of the critic-philosopher in the
name of truth is in fact a gorifcation of the poet as the prima
souce of this truth; if truth is the recogition of the systematic
character ofa certain knd of error, then it would be fully dependent
on the prior exstence of this error. Philosophers of science like
Bachelard or Wittgenstein are notoriously dependent on the aberra
tions of the pets. We are back at our unanswered question: does the
grammatiztion of rhetoric end up in negtive certainty or does i t,
like the rhetorization of grammar, remain suspended in the igno
rance of its ow truth or falsehood?
Two concluding remarks should sufce to answer the question.
First of all , it is not true that Poust's text can simply be reduced to
the mystifed assertion ( the superiorit of metaphor over metonyy)
that our reading deconstructs. The reading is not "our" reading,
si nce it uses only the linguistic elements provided by the text itself;
the distinction between author and reader is one of the false distinc
tions that the reading makes evident . The deconstruction is not
something we have added to the text but it constituted the text in the
frst place. A literar text simultaneously asserts and denies the au
thority of its own rhetorical moe, and by reading the text as we did
we were only trng to come closer to bing as rigorous a reader as
the author had to be in order to write the sentence in the frst place.
Poetic writing is the most advanced and rfned mode of deconstruc
tion; it may difer from critical or discursive witing in the economy
of its articulation, but not in knd.
But if we recognize the exstence of such a moment as constitu
tive of all literar language, we have sureptitiously reintroduced the
categories that this deconstruction was supposed to eliminate and
that have merely been displaced. We have, for example, displaced
the question of the self from the referent into the fgure of the
narrator, who then becomes the sigi of the passage. It becomes
18 RHETORI C
again possible t o ask such naive questions as what Proust's, or Mar
cel's, motives may have been in thus manipulating langage: was he
fooling himself, or was he represented as fooling himself and fooling
us into believng that fction and action are as easy to unite, by
reading, as the passage asserts? The pathos of the entire section,
which would have been more noticeable if the quotation had ben a
little mor extended, the constant vacillation of the narrator beteen
guilt and well-being, invtes such questions. They are absud ques
tions, of course, since the reconciliation of fact and fction occurs
itself as a mere assertion made in a text, and is thus productive of
more text at the moment when it asserts its decision to escape from
textual confnement. But even if we free ourselves of all false ques
tions of intent and rightflly reduce the narrator to the status of a
mere grammatical pronoun, without which the narrative could not
come into being, this subject remains endowed with a function that
is not grammatical but rhetorcal, in that it gives voice, so to speak,
to a grammatical syntagm. The term vie, even when used in a
grammatical terminolog as when we speak of the passive or inter
rogative voice, is, of course, a metaphor inferring by analog the
intent of the subject from the structure of the predicate. In the case of
the deconstructive discourse that we call literar, or rhetorical, or
poetic, this crates a distinctive complication illustrated by the
Proust passge. The reading reealed a frst paradox: the passge
valorizs metaphor as being the "right" litera fg, but then pro
ceeds to constitute itself by means of the epistemologcally incompat
ible fgure of metonymy. The critical discourse reveals the presence
of this delusion and afrms it as the irreversible mode of its truth. It
cannot pause there however. For if we then ask the obvous and
simple next question, whether the rhetorical moe of the text in
question is that of metaphor or metonymy, it is impossible to gve an
answer. Indivdual metaphors, such as the chiaroscuo efect or the
butterfy, are shown to b subrdnate fges in a general clause
whose sytax is metonymic; from this point of vew, it seems that the
rhetoric is superseded by a grammar that deconstrcts it. But this
metonymic clause has as its subject a voice whose relationship to this
clause is agin metaphorical. The narator who tels us about the
impossibility of metaphor is himself, or itself, a metaphor, the
metaphor of a grammatical syntag whose meaning is the denial of
metaphor stated, by antiphrasis, as its priorit. Ad this subject
metaphor is, in its turn, open to the knd of deconstruction to the
SEMI OLOGY AND RHETORI C 19
second degee, the rhetorical deconstruction of psycholinguistics, in
which the more advanced investigations of literature are presently
engaged, against considerable resistance.
We end up therefore, in the case of the rhetorical grammatiz
tion of semiolog, just as in the grammatical rhetoriztion of il
locutionar phrases, in the same state of suspended ignorance. Ay
question about the rhetorical mode of a literar text is always a
rhetorical question which does not even know whether it is really
questioning. The resulting pathos is an anxet (or bliss, depending
on one's momentar mood or indivdual temperament) of ignorance,
not an anxet of reference-as becomes thematically clear in
Proust's novel when reading is dramatized, in the relationship be
tween Marcel and Albertine, not as an emotive reaction to what
langage does, but as an emotive reaction to the impossibilit of
kowing what it might be up to. Literature as well as criticism-the
diference between them being delusive-is condemned (or
privleged) to b forever the most rigorous and, consequently, the
most unreliable langage in terms of which man names and trans
forms himself.
2 Topes
(Rlke)
RILKE I S ONE OF THE FEW POETS OF THE TWENTI ETH
centur to have reached a large and worldwde audience. Even in
France, where Yeats, Eliot, Wallace Stevens, Montale, Trakl, or Hof
mannsthal are not wdely kown, Rlke is more read than most of the
French poets of this centu. More than ff years afer his death, a
RiIe myth still lives wel beyond the borders of the Grman
speakng world.
The reasons for this degree of public prminence are not obvous,
for RiIe is not an easy or a popular poet. His work resists translation,
his themes are intimate, and his discoWse ofen oblique. Yet he has
been received wth a geat deal of feror, as if what he had to say was
of direct concern even to readers remote fom him in their language
and in their destinies. Many have read him as if he addressed the
most secluded parts of their selves, revealing depths the hardly
suspected or allowing them to share in ordeals he helped them to
understand and to overcome. Numerous biogaphies, reminiscences,
and letters bear witness to this highly personal mode of reception.
RiIke seems to be endowed with the healing power of those who open
up access to the hidden layers of oW consciousness or to a delicac
of emotion that refects, to those capable of prceivng its shades, the
rassuing image of their own solicitude. RiIe has himself ofen
played on the ambigit of a double-faced relationship toward
others, leaving in abeance which of the to, the poet or his reader,
depended on the other to nourish his strength. "I wish to help and
exect to b helped. Everone's eternal mista is to take me for a
healer when, in fact, 1 am ony attracting other, for my own proft, in
the trap of a simulated assistance." RIe confdes this self-insight in
connection wth a love afair, but it summarzes a mood encouraged
by some aspcts of the work. The initial seduction, the frst intimac
1. Letter from Rl to the Pincess of Thu and Taxis, Febr 2, 1915, in
Brhl Rk / Ma VN Thur un Ta (Ziich, 1951), 1:399.
20
TROPES ( RILKE) 21
between RIe and his readers almost inevitably occurs as an am
biguous complicit in shared confrontation wth "the near
impossibilit of livng.
,
,2 Sme passages of Malt, large fragments of
the corespondence, the general tonalit of Th Book of Hour, or a
somewhat hast reading of the Duino Elg-all orient the reading
in that direction. This tendenc, which RIe did nothing to discour
age, contributed much to the formation and the success of the per
sonal myth. It also lef extensive traces in RIe studies: it is some
times difcult to discoer the memor of the original texts under the
abundant confessional discouse that it generates in the commen
tators. RiIe's considerable audience is in part based on a relation
ship of complicit, on shared weakesses.
It is not difcult, for a reader alerted to the ambivalences of the
relationship between the self and its language, to demystif this
seduction. The intersubjective reading grounded in a common sen
timent, in the "transparenc of the heart," does not allow one to
reach the area of Rlke's poetr that is not afected by this demys
tifcation. In the case of this poet, readings that start out from the
most self-directed passages in the letters, the novels, or the confes
sional texts fail to uncover the potic dimension of the work. The
reason for this is not the bad faith which RiI confesses in the letter
from which I have just quoted; his poetr does not escape from
spathetic understanding because, under the guise of being solici
tous and disinterested, he does not hesitate, at times, to use others
rather coldly. The mechanics of this bad faith would b easy to
describe and, if the were indeed at the center of his consciousness,
the would be an efective way of access to his inner being. But the
are in fact peripheral and Secondar. It has not been difcult to call
into question the image of RIke as a healer of soul and to proe that
he was both less generous in practical and less stable in psychological
matters than one might have suspected.3 RiIe's intimate self re
mains in fact quite invisible and, far frm being its driving force, it
tends to vanish from the poetr altogether-which does not mean
that this poetr is deprived of a certain mode of inwardness that
remains to b defned. But the poet RiIe is less interested in his own
person than one might gather from his tone and from his pathos.
2. to des Lbns Fast-Unm6gicheit," same letter, p. 399.
3. S, for example, Peter Dmetz,R Rlk Pe Jar (Di sseldorf, 1953),
and Erich Simenauer, R M. mlk, Len un Myh (Bern, 1953).
22 RHETORI C
The nacissism that i s ofen ascribed to him no doubt exsts, but on a
ver diferent level from that of a reader using him as a refector for
his ow inner image. The personal seduction is certainly an impr
tant component of the work, but it functions, so to spak, as its zone
of maxmal opacit. One could approach and interpret a sizeable
part of his poetr by way of the negativ rod that would analyze this
seduction. It may be preferable however to tr to understand the
work in a less antithetical way and to read the poetic texts them
selves, rather than letters and confessional prose that may well turn
out to be of contingent importance.
On a somewhat more advanced level of understanding, the at
tractiveness of RIe stems from his themes. This is obvious, frst of
all, in the most supercial of ways: the potr puts on display a
brilliant variet of places, objects, and characters. As in Baudelaire,
the categories of the beautiful and the ugly are subsumed, in RI,
under the common rubric of the interesting. His poetic universe has
something dazzling, as if it consisted of rare items in a collection or a
museum, well set of against the background of a world that em
phasizes their singularit. Rpugnant and terrifng themes have the
same seductive power as the numberless objects of beaut and of
light-fountains, toys, cathedrals, cities of Spain and Italy, roses,
windows, orchards-that appar throughout the work. A form of
poetic decorum, itself a miture of caution and of genuine resere,
holds the volent images at a distance and prevents them from ac
quiring a presence strong enough to undo the fction or to dislocate
the language. No matter which of the uncanny fgures one singles
out, b it the epileptic in Malt, the stlite of the Ne Poem, or the
sinister acrobats of the Fifth Eleg, one will always encounter this
picturesque and surprising element mied with the horror and in
terposing, between the reader and the theme, the screen of a lan
gage that controls its own representational master. Even in what
appears to be Rlke's most personal poem, the poem wtten a few
days before his death and dealing wth his physical pain, the pain
remains "embellished" by the virtuosit of a prfectly prepared and
executed conceit.
It would be a mistake to dismiss this concern for attractive
suraces all too hastily as a form of aestheticism. The reference to
Baudelaire should sufce to stress that more is involved. Aesthetic
refnement is for RIke, as for the author of the Flur du ml, an
Apollonian strateg which allows him to state what would otherse
TROPES ( RILKE)
by unsayable. On this level of exprience, the aesthetics of beauty and
of ugliness can no longer be distinguished from each other. Nor is it
possible to thin of these seductive sufaces as merely supercial.
For the thematic attraction also functions on a more generally
inclusive level of understanding. Beyond the brightness of the set
tings, Rlke's work dares to afrm and to promise, as few others do, a
form of existential salvation that would ta place in and by means
of poetr. Few poets and thinkers of our centur have dared to go so
far in their afrmations, especially in afrmations that refuse to be
anchored in established philosophical or theological certainties, or to
h
ave recourse to ethical imperatives that might directly lead to
modes of action. It may seem surprising to characterize Rle's work
as psitive and afrmative when it puts such stress on the main
negative themes of modern consciousness. Rlke has an acute aware
ness of the alienated and factitious character of human reality, and
he goes far in his refusal to grant any experience the power to sus
pend this alienation. Neither love nor the imagining power of the
deepest nostalgias can overcome the essential barrenness of the self
and of the world. Severed forever from the plenitudes of self
presence, Rlke's fgure of humanity is the frailest and most exposed
creature imaginable. He calls man "the most ephemeral" (Ninth
Eleg), "the most fleeting" (Fifh Eleg), the creature "that is inces
santly departing" (Eighth Eleg), and that can never establish itself
in an appeased presence to itself or to the world. The promise that
the work contains is therefore anything but facile. But this maks it
all the more convncing.
On the thematic level, the existence of this promise is undeni
able. The large afrmations of the Elges, gnomic as the are, bear
wtness to this assertion, all the more so since the promise a salva
tion that could take place here and now: "Hiersein ist herrlich" ("To
be here is glorious" [Seventh Eleg]); "Hier ist des Sigih Zeit, hier
seine Heimat" ("Here is the time for the Tellabl, hre is its home"
[Ninth Eleg]); "Supernumerous existence / wells up in my heart"
(idem). This emphatic hr designates the poetic text itself and thus
afrms that it escapes the fragmentation of number and of time. In
the audacity of his assertion, Rlke assumes for poetr the furthest
reaching promise conceivable. The evolution of his own potr seems
to fulfl this promise. Afer being announced in the ElgeS, it comes
about in the appeased tonalit of the later work, the Snnt to
Orpheu, and many of the poems written afer 1912 and published
24
RHETORI C
posthumously . I t can be said of these poems that the perform the
transition from eleg to hymn, from complaint [Kge] to praise'
[Rhmn] .
One can understand therefore that Rlke not only claims the
right to state his ow salvation but to impose it , as it were, on others.
The imperative mode that ofen appears in his poetr ("You must
change your life"; "Demand change"; "Sing the world to the Angel"
.
.. ) is not only addressed to himself but asks for the acquiescence
of his reader. The exhortation is rooted in an authorit confrmed by
the possibility of its poetic exstence. Far from putting this assurance
in jeopardy, the insistence ofthe negative themes certifes its veracity.
A too easily ganted promise would be suspect and would not con
vnce, but a promise of salvation that could only be desered by
endless labor and sacrifce, in sufering, renunciation, and death, is a
diferent matter. One can begn to understand Rlke's poetr only if
one is wlling to entertain this conviction. As for deciding whether i t
i s a legitimate promise, whether it is a truth or a seduction, the
question must remain open, not only as a matter of caution but
because a rigorous reading must determine whether or not the work
itself ask this question.
The interreters who read Rlke's work as a radical summons to
transform our way of being in the world are therefore not misre
presenting him; such a summons is indeed a central theme of
the poetr. Some respond to it without reserations. Others have
suggested that Rlke is still in the gip of ontological presuppositions
which even the most extreme of his experences cannot reach and
that the reversal he demands, difcult as it may be, is still premature
and illUSOr. Rlke's good faith is not being questioned, but his blind
ness could be demonstrated by the critical analysis of his thought.
Heidegger had oriented the reading of Rlke in this direction, in an
essy published in 1949 which Rlkean studies have not yet entirely
assimilated.4 But it may b that the positivit of the thematic asser
tion is not entirel unambiguous and that Rlke's language, almost in
spite of its own assertions, puts it in quest ion.
This does not , at frst sight , seem to be the case. The advanced
level of refexive self-knowledge that informs Rilke's poetr nowhere
conflicts with the master of his poetic invention. The meaning ofthe
statement dovetails perfectly with the mode of expression, and since
4. "Wozu Dichter ... " in Holwege (Frankurt am Main, 1950).
TROPES ( RILKE) 25
this meaning possesses considerable philosophical depth, petr and
thought here seem to be united in a perfect synthesis.
For that reason, een the best interretations of RIe seem to
have remained, by and large, on the level of paraphrase, a para
phrase that is ofen subtle and carefl but that does not question the
convergence of the meaning wth the lingistic devices used to con
vey it. : The statements are rch enough in their content to sturate
the full range of meaning. The fact that these highly reflected state
ments directly implicate langage as a constitutive categor of mean
ing and thematize some of the lexcological and rhetorical aspects of
poetical diction by no means troubles the assumed convergence be
tween statement and T, between what is being said and the mode
of its sayng. RIe's propositions about langage are in fact caried
out in his poetr, thus allowing one to move freely between poetr
and poetics. The possibility of a confict between both never seems to
arise. Thus one of RIe's commentators can wite: "The poetic 'con
tent' and the poetic 'form' are so perfectly united in RIe's work that
it becomes impossible to object against the value of this poetr in the
name of a possible divergence beteen 'thought' and 'poetr e 6 Such
a divergence is inconceivable because RiIe is claimed to state, in and
through his poetr, the ver essence of poetr as the truth of this
essence. "The tre essence of poetr . . . is identical with the strc
tures of its poetic 'content'." In the author from whom we borow
these formulations, this trth is equated with an existential decision
that does not necessarily involve langage. But the exstential stance
must eventually lead to decisions that function on the level of the
language, even if these decisions appear to be of secondary impor
tance. The same commentator is natually led to consider formal
aspects of the poetr, such as rhyme or metaphor, but he at once
curbs their potential autonomy by fully identifing them with the
theme they convey: "The fundamental poetic practice, namely the
elaboration of a metaphorical language, also derives from the exeri
ence of sufering. The metaphor is an act of identifcation: the actual
sufering of the poet is made 'equal' with that of his symbolic
5. The studies of Rle that come closest to raising this question, wthout
however considerng it directly, are the work of Bd Aleman, Zit un Fir bim
spiten Rilk (Pfi llingen, 1961), and Maurice Blanchot's considerations on Rlke in
L'epae litrir (Pars, 1955).
6. Hermann Morchen, Rlks &mett an Orphu (Stuttgart, 1958), p. 20.
26 RHETORI C
fgures. . . ."
7
The ontological alienation that Rlke so eloquently
evokes would then not implicate language in any way. Lguage is
the unmediated expression of an unhappy consciousnss that it does
not cause. This implies that language is entiely ancillar in its rela
tion to a fundamental experience (the pain and the pathos of bing
which it merely refects, but that it is also entirely truthfl, since it
faithfully reprouces the truth of this pathos. The poet can thus
abandon himself without fear to his language, even to its most for
mal and outward features:
The logic of sounds [Lutlgk] to which the poet yelds when he
allows himself to be governed by the power of his language can
be meaningul only when it stands in the service of the truth
which this language uses in order to consere it. Poetr can be
truth only when its trust in language-a trust that is not
confned to acoustic afnities but that includes linguistic strc
tures in general, including etological relationship&-is in
deed attuned to this justifcation of existence which language, in
the region of its authentic origin, is always in the process of
formulating.8
With ver few exceptions, similar presuppositions underlie the
bst available critical readings of Rlke.9 One may well ask whether
the poetr indeed shares in the conception of language that is attrib
uted to it. Such a question difers entirely from a concentration on
the "form" of Rle's poetr, in the narrowly aesthetic sense of the
term; several careful studies have taen this approach but failed to
reach major exegetic results.10 By suggesting that the properly poetic
dimension of Rlke's work has been neglected in favor of his themes,
we do not wsh to return to the seduction of the forms. The question
is rather whether Rle's text turns back upon itself in a manner that
7. Morchen, p. 21; S also p. 15.
8. Morchen, p. 21.
9. The remark applies, wth qualifcations to complex to enumerate here, to
the witings on Rlke of Heideger, Guardini, Bollnow, Mason, and jacob Steiner.
10. Such as, for instance, H. W. Blmore, Rlk's Cranhip: An Angsi of
Hi Po
Stle (Oxford, 195); Ulrch Filebrn, D Stukturrble tr spitn
Lyk Rlk (Heidelberg, 1960); Frank H. Wo, Riner Mara Rlke: The Rn of
For (Minneapolis, 1958); Brigtte L. Bradle, Riner Mar Rlk neu Gcht:
Ihr zykliches Gge (Bern, 1968).
TROPES ( RI LKE) 27
puts the authority of its ow afrmations in doubt, especially when
these afrmations refer to the modes of writing that it advocates. At
a time when the philosophical interest ofRlke's thought has perhaps
somewhat declined, the present and future signifcation of his poetr
depends upon the answer to this question.
Rilke's work is ofen said to b divided b a clear brea that corre
sponds approxmately to the passage from Th Book ofHour to Th
Book of Imges ; it is also from this moment on that a degree of
master is achieved and that his manner reaches a certain stability.
+1
The brea marks an important modifcation in the metaphorical and
dramatic texture of the poetr. The more properly phonic elements
are less afected by it. Before and afer this date, Rle persists in
giving considerable importance to rhyme, assonance, and allitera
tion; in this respect, one can hardly speak of a major change, except
for a geater degree of refnement and control in the expressive use of
acoustic efects of language.
It is not easy to interpret this change. Commentators agree
neither on the meaning nor on the evaluation of Th Book ofHours,
and they have difculty locating it within the corpus of the complete
work. Certain characteristics of the situation and of the tone (a
prayer addessed to a transcendental entity) seem to prefgure the
Duino El; the volume also contains the frst mention of symbolic
objects and privileged words which will later acquire a central im
portance, whereas many of the other themes of Th Book ofHours
11. This bipartite division of the work does not correspond to a strct chronol
og in the composition of the poems. The texts that make up Parts I and II of Th
Bok of Hours (Th Bok of Plme and T Bok of Povr an Dath) date frm
Sptembr 1901 and from April 1903 rspectively, whereas some of the texts in
cluded in Th Bok of Im go back as far aSJuly 1899 and thus at times antedate
even Th Book of Monast Li which was wtten in September 1899. Yet takng the
manner and the stle of the subsequent work as a norm, Th Book of Imge
certainly appears more "advanced" than Th Bk of Hour. Rther than a genetic
devlopment, we are dealing with to distinct potic manners that can est side by
side. According to the same stlistic critera, Th Uf of Mar, which dates from
1912, would belong to Rlk's youthful work. This provs that the distinction beteen
what is called "early and "late" work is ofen not as simple as a genetic terminolog
would lead one to believe. Still, the break to which we U here alluding ofers a
convenient point of reference for the organiztion of the work. A long as one does
not confer upon it the power that belongs to origins, the division has a certain
validit.
28
RHETORI C
disappear from the later work. 12 The feror with which the poems
address a power that is given the name of "God" raises the question
of their theocentric strcture, a question that never stops haunting
the exegesis of Rlke wthout, however, receivng a satisfactor an
swer. 13 Like iron flings under the power of a magnet , the verbal
mass tuns towards a single object that causes the eclosion of an
abundant potic discourse. The following pom, a tpical instance of
Rlke's poetr at this time, can both give us some notion of this
discourse and sere as an introduction to the general problematics of
the work. Since we have to allude to sound elements that cannot be
translated, I quote in German:
Ich liebe dich, du sanfestes Gesetz,
an dem wir reifen, da wir mit ihm rangen
du grosses Heimweh, das wir nicht bzwangen,
du Wald, aus dem wr nie hinausgegngen,
du Lied, das wir mit jedem Schweigen sangen,
du dunkes Netz,
darin sich fichtend die Gefthle fangen.
Du hast dich so unendlich goss begonnen
an jenem Tage, da du uns begannst,
und wr sind so gereif in deinen Sonnen,
so breit geworden and so tief gepfanzt,
dass du in Menschen, Engeln und Madonnen
dich ruhend jetzt vollenden kannst.
Lass deine Hand am Hang der Himmel ruhn
und dulde stumm, was wir dir dunkel tun.
[1:24]
12. As in the fgure of the ball, metaphor for G and addressed as "Ding der
Dinge" in a poem from Th Bok ofMoTti Li (Rnr Mari mlke, Werke in dre
Bin [Frankfurt am Main: Insel Verlag, 196], 1:21. All quotations from Rlke's
pms refer to this edition). The passge can b read as an early vrsion of some of
the Ne Poem.
13. Else Buddeberg, Rin Mari mlke, ei innr Biaphi (Stuttgart,
19
5), obsere, wth refernce to Th B k o Hour, "that the ealution of [its]
mert is still being much discussd toa and quotes two crtics to illustrate the
wide divrgences of opinion. One of them spa of a "relinquishing of the self to
G + . . as the Grman language had not kow since H6lderlin," whereas the
other asserts that it would be a "sentimental confusion ... to fnd the slightest trace
of serious religious feeling in this book" (Buddeberg, p. 531).
TROPES ( RI LKE) 29
By its setting, which follows the convention of the ode as a series
of reiterated apostrophes that are as many metaphors, the poem
indeed seems to be fully centered on the entit it attempts to name.
But the periphrastic designation is so divrse that it becomes vage:
the entt is addressed as "law," "homesickness," "forest," "song,"
and "net," a sequence that cannot easily b reduced to a common
denominator. Moreover, the entit is never itself desigated by one of
the attributes that proprly belong to it. The play of personal pro
nouns is balanced beteen "I" (or "we") and ''you,'' thus establishing
a nearly perfect symmetr from which the third person is practically
excluded; afer the "ihm" in the secOd line, the "ich/du" or "du/wir"
pattern is close to perfect.14 The object of the apostrophe is only
addressed in terms of an activit that it provoks in the addressing
subject: if it is said to be a forest, it is only with reference to our
behavior towards this forest; the net exists only as an obstacle to our
fight; law is, per defnition, that which governs our behavor and the
song is at once identifed as our song (or silence) . The metaphors
therefore do not connote objects, sensations, or qualities of objects
(there is practically no third person in the grammar of the poem1S),
but refer to an activt of the speaking subject. The dominating
14. The difcult of translation, especially in the earlier Rle poms, rmain
vsible in the vrsion produced b J. B. Leishman (Riner Maria Rlk, Sl Work
[London, 196], p. 3);
I love you, gentlest law, through which we yet
were ripening while with it we contended,
you great homesickess we have not transcended,
you forest out of which we never wended,
yu song that from our silence has ascended,
you somber net
where feelings takng fight are apprehended.
You made yourself a so immense bginning
the dy when you bgan us to,-nd we
bneath your suns such ripenes hav ben winning
have gow b broadly and deep-rotedly,
that you, in angels, men, madonnas inning,
can now complete yourslf quite tranquilly.
Let your right hand on heaven's slop repose
and mutely bar what drky impos.
15. An exception occurs in line 7, where the third persn [d Ghl] precisely
refers to the feeling, to the interorit of the subject.
30
RHETORI C
center, the "du" ofthe poem, is present in the poem only t o delegate,
so to speak, its Fotential activit to the speang voice; this becomes
the explicit theme of the poem in the two concluding lines. The
purpose of the text is not to reunite the two separate entities but to
evoke a specifc activit that circulates between them.
The poem does not mention this activit by name. It states
instead, in its fal sentence, that it must remain obscure and invisi
ble: "dunkel tun. " That it is called a flfllment [Vollnung] and
that the wll of the "du" is said to be accomplished by this act does
not allow for its defnition but repeats in fact the relationship of
immanence between the two "persons" that is being staged in the
text . A more implicit reading permits however some further speci
fcation. The beginning of the poem indicates that the activit in
question is frst perceived as a constraint and provokes the vain
attempts to escape from its power. This is being openly stated in the
frst two lines and more sugestively evoked in the two following
ones: the homesickness is oppressive, but we cannot evade it ; there
can be no escape from the forest that surrounds us; silence itself
cannot prevent us from singing. The sequence culminates in the
fge of the net : feelings that tr to escape into forgetting or into
indetermination are imprisoned and coerced, by this activt, to re
main present to us.
But the constraint changes to acquiescence. In the second
stanza, the relationship beteen the "I" and the "you," instead of
being paradoxcal and dialectical as in the frst section, blossoms out
in the luminous image of the tree. The promise of the beginning
fulflls itself as naturally and harmoniously as the ripening of fruit in
the sunshine. The transformation designates the acquisition of a
greater master in the activit that the poem symbolizes. This mas
ter is thematically asserted in the reversal that has takn place
between the beginning and the end ofthe poem: the subject that was
at fst compelled to obey can now act in full freedom and can
conform its will to that ofthe "law. " The central wll ofthe poem has
been transformed from constraint into a benevolent sun, wth only
the
repetition of the wor "dark" (dunles Netz, dunkel tun) as a
reminder of the original violence. Besides, the mention of "hand" in
the next-to-Iast line strengthens the impression that we are dealing
with an action involving skills that the initially reluctant student now
fully masters.
The proof of this master can only be hidden in the text. The
TROPES ( RILKE)
31
relationship between the two subjects or grammatical "persons" is so
tight that it leaves no room for any other system of relationships. It is
their interlacing that constitutes the text. There is therefore nothing
in the poem that would entitle us to escape beyond its boundaries in
search of evdence that would not be part of it: the freedom that is
afrmed at the end is precisely a freedom wthin bondage that can
prevail only because it is tolerated by the authorit of the power
which allows it to exst. It remains subjected to the single authorit,
to the single achievement, of the text.
This achievement, however, is prmarily phonic in kind. The last
stanza, in which the master is asserted, is also the one in which
efects of euphony reach their highest point of elaboration. The poem
comes to rest in the lines
Lass deine Hand am Hang der Himmel ruhn
und dulde stumm, was wr dir dunkel tun.
It can easily be verifed that, in this last line of verse, there appears
rigorously no syllable that does not fulfll an efect of euphony. The
main rhyes and assonances (dulde stumm, wir dir, dunkel tun) are
interconnected by sllables that are themselves assonant (und dulde)
or alliterated (was wir) and thus enclose each sound-efect into
another, as a larger box can enclose in its turn a smaller one. The
master of the poem consists in its control over the phonic dimen
sions of language. A reading of the other poems in Th Book of
Monti Li conrms this conclusion. The "G" that the poems
circumscribe by a multitude of metaphors and changing stances
corresponds to the ease that the poet has achieved in his techniques
of rhyme and of assonance. It is well known that these poems were
written ver quicky in a kind of euphoria which Rilke wll remember
when, more than twent years later, he wll write the &mnt to
Orphu; what the poems celebrate is primarily this euphoria. The
metaphors connote in fact a formal potential of the sigifer. The
referent of the poem is an attribute of their language, in itself devoid
of semantic depth; the meaning of the poems is the conquest of the
technical sklls which they illustrate by their acoustic success.
It may seem preposterous to associate such a near-mechanical
procedure wth the name of Gd. Yet, the apparent blasphemy can
just as well be considered as the hyerbole of an absolute phonoen
trism. A poem of Th Book ofMonti Li (1 :20) asserts the possi-
RHETORI C
bilit of overcoming death itself by means of euphony, and i t fulflls
this prophec in its own texture, in the "dark interal" [im dunkln
lntrall ] that in its assonance both separates and unites the two
words ''Tod'' (death) and ''Ton'' (sound). Once we succeed in hear
ing the song hidden in language, it will conduct us by itself to the
reconciliation of time and exstence. This is indeed the extrvagant
claim made by these pems when the pretend to designate G by
means of a medium which deprives itself of all resouces except
those of sound. Possibilities of representation and of exression are
eliminated in an askesis which tolerates no other referent than the
formal attributes of the vehicle. Since sound is the only property of
language that is trly immanent to it and that bears no relation to
anything that would be situated outside language itself, it will re
main as the only available resource. The Cratlic illusion, which is
held by some to constitute the essence of poetr and which subordi
nates the semantic function of language to the phonic one, is
doubtlessly at work in Th Book ofMont Li. In a manner that is
not yet entirely convincing, this early volume already partakes of the
Orphic myth.
In these texts, in which a measure of technical master alter
nates with moments of clumsiness, the failure of the claim is as
evident as is its presence. In order to give a coherent framework to
the sequence of poems, RIke is forced to substitute a subject that tells
the stor of its experience for the unmediated beauty of the petic
sund. The poems thus acquire a meaning that does not entirely
coincide wth their actual intent . The introduce an autonomous
subject that moves in the forefront and reduces the euphony to the
function of orament. In the frst version of Th Book ofMonat
Li, this impression was still heightened by the brief narrative sec
tions inserted between the poems, like a journal commenting upon
the daily progress of the poet's work.16 The fact that RiIke was
obliged to invent a fctional character, a monk surrounded by his
ritualistic paraphernalia, well illustrates his inability, at that time, to
dispense wth the conventional props of poetic narration. Ad since
the subject is confned to being an artisan of euphony, it has only a
16. This frst version of Th Bk i Monti Li appeas uder the title Di
Gb in the third volume of the Cmpl Wor published in si volumes by Ernst
Zinn C&tlih Werke [Frankrt am Main: Insel Verlag, 1955-66], 3:305-73).
TROPES ( RILKE) 3
rather thin stor to tell. In the to subsequent volumes of Th Book
ofHour, especially in Th Book ofPovr an ofDath
,
RiIe aban
dons the claim to a self-referential diction and returns to the direct
expression of his ow subjectivit. The texts lose most of thei formal
rigor and acquire the obvous interest of a self-narrating sensibilit.
These poems are easy of access and ofen moving, but measured b
RIe's fnal and initial ambition the represent the least exalted
moment of his poetic production. It wll take the long labors ofMalt
and of Th Ne Poem to reconquer the impersonalit that was
proclaimed and lost in Th Book ofth Mont Li.
Wile he was witing Th Book ofHour, RiIe was also working at a
ver diferent kind of poem that would fnd a place in Th Book of
Imge, itself a work of transition leading up to the masterful Ne
Poem. The development that taes place in these texts is decisive for
the entire mature work. It can be described by the reading of one of
the poems characteristic of this period. The poem entitled "A
Rnde der Nacht" ("At the borderline of the night") is a somewhat
arbitrarily chosen but tpical instance:
Am Rnde der Nacht
Meine Stube und diese Weite,
wach iber nachtendem Land,
ist Eines. Ich bin eine Saite,
iber rauschende breite
Rsonanzen gespannt.
Die Dinge sind Geigenleiber,
von mu endem Dunel voll;
drin traumt das Weinen der Weiber,
drin rihrt sich im Schlafe der Groll
ganzer Geschlechter . . .
Ich solI
silbem erzittern: dann wrd
Alles unter mir leben,
und was in den Dingen irrt,
wrd nach dem Lichte streben,
das von meinem tanzenden Tone,
ur welchen der Himmel wellt,
3
durch schmale, schmachtende Spalten
in die alten
Abgi nde ohne
Ende fcIt . . .
[ 1: 156]
1
7
RHETORI C
Instead of being caught in the "somber net" of a pseudo
dialectic beteen pseudo-subjects, we are at once within a much
more familiar ptic landscape. From the beginning, the poem an
nounces itself as naming the unit, the complementarit of an in
side/outside polarit: the inner seclusion of the "room" (which intro
duces a subject by the possessive of "my" room) and the infnitely
wide epanse of the nigt outside. The are decreed to be on by
categorical assertion, as if this unit were the sudden revelation ofa
single moment, a specifc accord between the self of the pot and the
world that surrounds him. But the poem does not remain within the
instantaneous stasis of this accord. The initial oneness undergoes a
transformation announced in lines 1 1 and 12: "Ich solI / silbern
erzittern. " This event trigers a transformation which is exerienced
as a movement of expansion. It is no longer the static unity ofinside
and outside that is being asserted, but the metamorphosis of an
oppressive and constraining inwardness into a liberating outside
world. The positive valorization of the movement is marked by the
ascending motion of darkness towards light : " . . . was in den
Dingen irrt , / wird nach dem Lichte streben. " Upon the schronic
axs of an inside/outside polarity is juxtaposed a dynamic axis which
transforms the inside/outside opposition into a successive polarity of
the te night/day.
For a reader accustomed to Rmantic and post-Rmantic
poetr, this tpe of poem is most familiar, both by what it asserts
and by the antithetical couples that it sets into play. It tries to evoke
and accomplish the sythesis, the unity ofa consciousness and of its
objects, by means of an expressive act , directed from inside to out-
17. Translated as literally as the text allows: "My room and thi wide space I
watching over the nigt of t he land-I are one. I am a string I strung over wide,
r
o
ring r
esonances. II Things ar hollow violins I full of a goanng dark; I the
laments of women I the ire of generations I dream and toss within . . . I I must
t
r
mble I and sing lie silver: then I Al will live under me, I and what ers in things I
will strive for the light I that, from my dncing song, I under the cwe of the sky I
throug langishing narow clefs I falls I in the ancient depths I wthout end . .
TROPES ( RI LKE)
35
side, which fulflls and seals this unity. The subject/object polarity,
which remained vague and ambivalent at the beginning, is clearly
designated when the poem explicitly confronts the subject, no longer
with the indefnite immensit of the frst line, but with the objects,
the particular thing that are contained in this wide space. The unit,
which was only asserted as a prr at the start, actually occurs
before our ees when the subject, claiming to be the string of a violin,
meets and adapts itself perectly to objects which, in a metaphor that
is trly Rlkean in its seductive audacity, are said to be the "body" of
this same violin, "Gigenleiber." The totality of the One thus consists
of a perfect complementarity: without the sounding board of the
volin, the string is devoid of value, but it sufces to bring them
together to mae the "somber and deep unit" of the world vbrate
and shine. Everthing seems to confrm that this poem can be con
sidered a later version of the "correspondence" beteen the inward
ness of the subject and the outside world. The exteriorit is further
confrmed by the assimilation of the sky's immensit, in the frst line,
to a thing; it is indeed the resonance of its space ("Ich bin eine Saite, /
tber rauschende breite Resonanzen gespannt") which is transformed
in the musical by of things ("Die Dinge sind Gigenleiber"). The
pom is an exmple of the most classical of metaphors, conceived as
a transfer from an inside to an outside space (or vce versa) by means
of an analogical representation. This transfer then reveals a totaliz
ing oneness that was originally hidden but which is fully revealed as
soon as it is named and maintained in the fgural language. One
could stop here, and confne oneself to the discover of frther
analogical parallels (such as the convergence of the spatial with the
musical theme by way of an erotic connotation-since the body of
the violin is that of a woman as well) and especially by stressing the
perfect coalescence of the metaphorical narration with the sound
pattern of the poem. The moment of synthesis corresponds exactly to
the modulation of the assonances from the i sound (ten times re
peated in the frst eight lines) to the e sound (ten times repeated in
the four last ones) . One should also draw attention to the detailed
precision in RiIe's selection of metaphorical analogons.
But if one allows oneself to be guided by the rigorous repre
sentational logic of the metaphors, whose clarity of outline indeed
distinguishes a poem like this one from those of Th Bok ofHours,
then one should follow their guidance to the end. For RIe's singular
ity becomes manifest in a displacement that distorts the habitual
36
RHETORI C
relationship between theme and fgre. The pattern we have just
schematized dos not appear quite in this shape in the text . The
inwardess that should belong, per defnition, to the subject is lo
cated instead within things. Instead of being opaque and full, things
are hollow and contain, as in a box, the dark mass of sentiments and
of histor. The interiorit of the speakng subject is not actively en
gged; whatever pathos is mentioned refers to the sufering of others:
the woes of women, the ire of historical generations. By a cuious
reversal , this subjectivit is invested from the start, before the fgral
trasfer has taken place, in oects and in things. This subjctiv e
perence is sid to b dark to the etent that it is uable, by itslf, to
fnd exression; it exsts in a condition of error and of blindness
("was in den Dingen iI . . . ") until the subject, the "I" of the poem,
confers upon it the clarit of entities that are available to the senses
by gving it the attribute of voice. The usual structue has been
reversed: the outside of things has become internalized and it is the
subject that enables them access to a certain form of exteriorit. The
"I" of the poem contributes nothing of its ow exprence, sensa
tions, suferings, or consciousness. The initial moel of the scene is
not, as one might thin at frst, that of an autonomous subject
confronting nature or objects, as is the case, for example, in
Baudelaire's poem "l'Homme et la mer." The assimilation of the
subject to space (as the string of a violin) does not really occur as the
result of an analogical echange, but by a radical appropriation
which in fact implies the loss, the disappearance of the subject as
subject. It loses the indivdualit of a particular voice by becoming
neither more nor less than the voice of thing, as if the central point of
view had been displaced into outer things from the self By the same
token, these outer things lose their solidity and become as empty and
as vulnerable as we are ourselves . Yet, this loss of the subject's au
tonomy and of the resilience of the natural world is treated as if it
were a positive event, as a passage from darkness to light. It would
be mistaken to interpret this light a the clarty of a self-knowledge.
In the logc of the fgure, it is nothing of the sort: the light is the
transformation of a condition of confusion and of non-awareness
(dream, sleep, erring) in the sound-version of this same, unchanged
condition. The fgure is a metaphor of a becoming-sound, not of a
becoming-conscious. The title, "At the borderline of the night,"
should not be read as the daw of a new lucidity but rather as a
persistent condition of confusion and dispersion from which there is
TROPES ( RI LKE)
37
no escape. The end of the poem conrms this reading: the rising light
tuns out to be a fall in "the ancient / depth without / end . . ." of
the night. The totalization taes place by a retun to the emptiness
and the lack of identit that resides in the heat of thing. The unit
afmed at the begning of "A Rnde der Nacht" is a negtive
uit which deprives the self of any illusion of self-insight. By becom
ing a musical string, the self partakes forever in the erring of thing.
Yet, it gves voice to this eranc.
This reversal of the fgal order, itself the fgre of chiasmus
that crosses the attributes of inside and outside and leads to the
annihilation of the conscious subject, bends the themes and the
rhetoric from their apparently traditional mode towards a specif
ically Rlkean one. It is difcult to comprehend this reversal on the
level of the themes. The notion of objects as containers of a subjectiv
it which is not that of the self that considers them is incomprehen
sible as long as one tries to understand it from the perspective of the
subject. Instead of conceiving of the poem's rhetoric as the instru
ment of the subject, of the object, or of the relationship between
them, it is preferable to reverse the perspective and to conceiv of
these categories as standing in the serce of the langage that has
produced them. The metaphor of the violin fts the dramatic action
of the text so perfectly and the image seems so fawlessly right be
cause its external strcture (box, string, clef that produces and lib
erates the sound) trigers and orders the entire fgal play that
articulates the poem. The metaphorical entit is not selected because
it corresponds analogically to the inner experience of a subject but
because its strcture corresponds to that of a lingistic fgre: the
violin is lik a metaphor because it transforms an interior content
into an outward; sonorous "thing." The openings in the box (so
fttingy shaped like the algorithm of the integal calculus of totaliza
tion) correspond precisely to the outside-directed tun that occurs in
all metaphorical representations. The musical instrment does not
represent the subjectivit of a consciousness but a potential inherent
in langage; it is the metaphor of a metaphor. What appears to be
the inwardess of things, the hollow inside of the box, is not a
substantial analog between the self and world of things but a for
mal and strctural analog between these' things and the fgral
resources of words. The coming into being of metaphor corresponds
point b point to the apparent description of the oect. But it is not
surprising that, in evoking the details of the metaphorical instrment
38 RHETORI C
or vehicle (the perect ft of the sting to the box, the openings in the
sounding-board, etc. ), the metaphor comes into being before our
ees, since the object has been chosen exactly for this purpose. T
h
e
corespondence does not conm a hidden unity that exsts in the
nature of things and of entities; it is rather lie the seamless encase
ment of the pieces in a puzzle. Perect adjustment can take place
only because the totality was established beforehand and in an en
tirely formal manner.
The poem "Am Rnde der Nacht" still disguises this strateg by
siuatig the bith of metaphor as the confation and the proof
of the unity apodictically announced at the beginning ofthe text. But
a caeful reading can reveal the stratagem without having recourse to
outside information. The poem, which fst appeared to be a con
frontation betwen man and nature, is in fact the simulacrum of a
description in which the strcture of the described object is that of a
fgral potential of langage. Moreover, one should not forget that
the metaphor of the metaphor is represented as an acoustical pro
cess: the metaphorical object is, literally, a musical instrument. The
pe encasing of the fgs ms language sing li a violin. The
priorit of the phonic element that was stressed with regard to
Th Book ofMonati Li has not been abandoned. Not only is it
audible in the parallel between the symbolic action and the euphony
of the assonances, but it extends to the play offgration. Th Book oj
Images is not less "phonocentric" than Th Book oJHours-far from
it, since now the imperatives of euphony govern not only the choice of
words but the choice of fgures as well .
The lingistic strateg of this still relatively ealy poem (which has
several equivalences among the other texts that make up Th Book oj
Imge) wll dominate the work until the end. The determining
fgre of Rle's poetr is that of chiasmus, the crossing that reverses
the attributes of words and of things. The poem are compsed of
entities, objects and subjects, who themselves behave lie words,
which "play" at language according to the rules of rhetoric as one
plays ball according to the rles ofthe game. "Am Rnde der Nacht"
is particularly revealing because it still makes use of the classical
schema of a subject/object dialectic. The lingistic chaacter of one
of the poles involved in the inversion is therefore relatively easy to
perceive, whereas it will ofen be hidden in the later work. At the
same time, the almost programmatic tonality of the poem, the unit
TROPES ( RI LKE)
39
frst asserted and then "demonstrated" b the transformations of the
fgres , will also disappear. In the Ne Poem (Neu Gdht) the
same poem would have been constructed diferently. It might have
been called "The Violin"; the two frst lines would in all probability
have been replaced by a description that reverses the "real" schema
of events: instead of being the result of their union, it might have
been music itself that brought the string and the violin in contact
wth each other. A poem li the following, the entrance text to Th
Book ofImge ("Eingng," 1:127) , clearly indicates the structure of
the reversal . In the evocation of what could be called an abridged
landscap, the reversal appears in the fact that the ees of the person
who is being addressed constitute a world of objects, instead of the
objects directing their glance:
Mit deinen Augen . . .
hebst du ganz langsam einen schwarzen Baum
und stellst ihn vor den Himmel: schlank, allein.
Und hast die Welt gemacht. 18
The world which is thus created is then explicitly designated as a
verbal world. Contact with this world is comparable to the discover
of meaning in an interpretation, and the interpretation engenders
the text by appearing to describe the object:
Und hast die Welt gemacht . Und sie ist gross
und wie ein Wort, das noch im Schweigen reift.
Und wie dein Wille ihren Sinn begreif,
lassen sie deine Augen zartlich los . . . 1 9
But this poem is something of an exception. In the vast majority
of the Ne Po, only the structure of reversal is maintained, and
its orientation towards the pole of language remains implicit . This
remark gives access to the dominant pattern of the mature work, but
it also implies the possibility of a misreading which will become an
integal part of the poetr till well into its latest developments.
18. "With your eyes . . . / You slowly lif up a black tree / and stand i t, thin,
alone, before the sky. / You made the world. "
19. "You made the world. And it i wide / and like a word that ripns still in
quet / And once you vouch to understand their sense / They'll gently let your eyes go
free . . .
40 RHETORI C
By showing the prevalence, in the Ne Poem, of this reversal ,
one can also isolate the poles around which the rotation of the
chiasma takes place. A is clear from the titles of the indvidual
pems that mae up the Ne Pom, the are ofen centered on
natural or man-made objects. When the describe personages or
settings, the have ofen ben so caught in a stlized perception that
the have become like icons, emblems of a feeling or of a destiny as
sharply circumscribed as are the properties of things. It soon appears
that all these objects share a similar fndamental strcture: the are
conceived in such a way as to allow a reversal of their categorical
properties, and this reversal enables the reader to conceive of proper
ties that would normally be incompatible (such as inside/outside,
bfore/afer, death/life, fctionrealit, silence/sound) as complemen
tar. The engender an entit, like the violin and the string of "A
Rnde der Nacht, " which is also a closed totality. Ifw question why
such or such an object inscribed in the Ne Poe has compellingy
attracted RIe's attention (or why he deliberately selected it) , the
answer will always be that it forced itself upon him because its
attriutes alow for such a reers and for such an (apparent) totali
ztion.
A particularly clear and concrete instance of such a structural
reversal would be, for example, the specular refection. The poem
"Quai du Rosaire" ( 1 :290) is a fne case in point. Taking advantage of
a ligt efect at dusk, Rlke can, wthout seeming to be fantastic,
decree that the upside-down world that is refected in the still water
of the canals is more substantial and more real than the ordinar
world of the day:
das abendare Wasser . . .
darin . . .
die eingehangte Welt von Spiegelbildern
so wirklich wrd wie diese Dinge nie.
20
The description of the details of this upside-dow cit, althoug
it maintains the realism of the local color (Estaminets, 1 . 16) one
expects in a poem that is also like a postcard, thus acquires a some
what uncanny and as it were sureal character. The reversal of the
20. "the clear evening water . . / in which + . . / the suspended worl d of
mirored images / becomes more real than things ever were. "
TROPES ( RI LKE) 41
attribute of reality (the text stresses indeed reality, "Wirkich[keit]")
was prepared from the frst part on. In an apparent personifcation,
which is in fact a prosopopoia based on the laguage-embedded
idiom accordig to which, in Grman as in English, streets ae sid
to "go" from here to there, the aulia condition for an action (the
streets, auxiliar device for the action of going) bcomes the agent of
this same action. The slight note of absurdity sounded in the fst
evocation of the walng streets ("Die Gassen haben einen schten
Gang / . . . und die an Platze kommen, warten lang / auf eine
andre, die mit einem Schrtt / iber das abendkare Wasser tritt
. . . ")2 1 prefgures the reversal of the refection which might other
wse seem too brusque or artifcial .
The sureality is not limited to the refected world. We sw that
the reversal acquires poetic value only when it leads to a new totali
zation; this is why, after having traversed the surface of the lookng
glass and entered the refected world, the poem has to retun, in the
last stanza, to the real world "above. " By the same token, the tem
poral natue of an event that , up till then, was descrbed in spatial
and ocular terms, becomes manifest. The blurring of the outlines,
which at frst seems to be due entirely to the play of light and
shadow, taes on a temporal dimension when one remembers that
the poem is about "Brugge, " "Brges la morte" as it is called by the
poet Gorges Rodenbach, a cit that used to be prestigous but has
bcome, by the loss of its natual harbor and medieval gor, an
emblem for the transience of human achievement , a fge of muta
bility. The question that introduced the temporal dimension, "Ver
gng nicht diese Stadt? ("Did not this cit prish?") , a question riter
ated in line 17: "Und oben blieb?" ("Ad what remained above? ) , is
answered at the end: the real world "above" has not been entirely
dissolved in the refection of things past , since the fnal perception
(the bells of the carillon) reach us from abve. But this realit is then
no longer solidly anchored on the ground. The refection has emptied
it out ; its illusr stabilit has been replaced by the sural irrealit of
the mirror image. The descent in the underworld of the mirror
uplifs the real and suspends it in the sky, like a constellation. The
fnal totalization takes place within this constellation, which could
21. "The streets go with a gentle walk / . . . and when the reach the suares
they wait / forever for another which, in one sole step / crosses the clear evning
water . . .
42
RHETORI C
not have come about wthout the passage through the fction of the
specular world.
This new totalit is itself temporal in kind: the sound of the
ca
rillon, the real totalit tht rmin, also has for its fnction to
measue the passage of time. By thinkng of Brge no longer as a
stable reality but as the fge of temporal loss and erosion, the
reality lost in the everday world ofunrefected surfaces is recovered:
the live Brge is much less "r" than "Brges la morte."22 Fa y,
the temporal constellation that fnctions as a resolution manifests
itself, in the last analysis, as soun. Perceived in the trth of its
mutabilit, time becomes an audible reality.
This eperience oftime is highly paradoxcal . It acquiesces to all
that ordinarily appars as the oppsite of permanence and of dura
tion. The afrmation is retained in the seductive but funereal image
of a temporal annihilation which is enjoyed as if it were a sensuous
pleasure, "der Sussen Traube / des Glockenspiels" ("The sweetened
cluster of grapes / of the carillon") , which actually is the death knell
that reduced the city to a ghostly memor. Similarly, the sound of
this new temporality will have all the attributes ofits opposite: at the
end of the pom, a new chiasmus crosses the attributes of silence
and of sound and. designates the sound of the carllon by the proper
ties of silence:
Und oben blieb? -Die Stille nu, ich glaube,
und kostet langsam und von nichts gedrangt
Beere ur Beere aus der sussen Traub
des Glockenspiels, das in den Himmeln hangt . 23
22. Rle himself sys just abut exactl the opposite in a prose tet entitled
Fm (Werk in dri Brn, 3:498) , which begins with considerations on the city
ofBrge. The prose text hardly invalidtes ou reading. Al it proves i that this text ,
which is a knd of travel joural, does not sy the same thing on a gven entity ( the
city of Brge, in this cas) as the ptr. The passage is a goo exmple of the
dnger inherent in a too literal use made of the "soues" derived from the prose
work or from the letters. The specifc moment that Rle wihed to retain for the
pm
a appar i the pros pa ge: "It i cnstantly vanshing, like a f eaten
by the lacework of dampness . o" ( 3:498) . Fum also contain the exlicit refer
ence to
Brge l Mor by the Femish sbolist poet (who wte in French)
Gorges Renbach.
23. "Ad what remaied abve?-ny silence, I beliee, / which tastes slowly
and unhuriedy / gape b grape the sweetened cluster / of the carlon, suspnded
in the skes. "
TROPES ( RI LKE)
43
The evoation of Bruge as the image of mutability is in itself
banal ; if it were to b reduced to this theme alone, the poem would
be of minor interest . The recover of duration by means of the
subject's acquiescence to the temporal erosion that threatens it is
more challenging: it combines the audacity of a paradox with a
promise of beauty or even, in the image of the grapes, of sensuous
gratifcation on the far side of the gave. Yet the true interest of the
poem does not stem from these thematic statements, but rather
from the intricac and the wealth of movements trigered by the
original chiasmus. The crossing of the categories of reality and of
specular refection articulates a sequence of similarly strctured re
versals: reversal of agent and instrument , of ascent and descent , of
inside and outside, ofloss and recuperation, death and life, time and
sound, sound and silence. A great deal of rhetorical agitation is
contained in a brief petic text which also has the innocent appar
ance of a picturesque description, of a picture postcard.
Versions of this same patter reappear in each of the Ne
Poe. Each of these poems is closed of in its own self-sufcienc as
the description of a particular object or scene, and each poem states
in its own terms the enigma of the chiasmus that constitutes i t .
"L'ange du meridien," for example (to refer only t o the best known of
the Ne Poem, 1 :253) , culminates in the totalization of a temporal
ity which can, in opposition to the lacunar time of everday experi
ence, be said to be full ; this total time is evoked by means of the
fgur of a sundial which, during the night, registers time that would
b as entirely imaginar as might be invisible light . The temporal
totaliztion is brought about by the chiasmic reversal of the
categories night/day and light/dark. "Or Ball" ( 1 :395) is a strictly
descriptive version of a totalization that includes the contradictor
motions of rising and falling [Flug un Fall ] . 24 It is brought about by
means of an object which, like the violin in "A Rnde der Nacht, "
has become the depositor of an inwardness which is not simply that
of the subject . 25 The moment of reversal is graphically represented
2. The totalizton of rise and fall is one of the fundamental trops of Ril's
poetr. It is themtically assrted at the end of the Tenth Eleg but recurs persis
tently throughout the work. "Das Kapital" ( 1 :257) would be a characterstic instance
among other. The theme is present in Th Bok of Monti Li, although it would
b prmature to spak of totaliztion in this cas.
25. It goes without sayng that thi movement , which occurs in the lines " . . .
was in den Ggenstanden / nicht bleiben kann , / das gitt in dich . . . " {"what
RHETORI C
when the subject becomes, in its turn, a thing whose motion is
determined by another thing at the precise instant when the ball
reaches the apogee of its own trajector:
[der Ball] . . . und sich neigt
und einhalt und den Spielenden von oben
auf einmal eine neue Stelle zeigt ,
sie ordnend wie zu einer Tanzfgur,
_ _ b
The reversal makes it possible to consider the falling motion as if it
were an event that partakes, to some degee, in the joyful upsurge of
the ball's frst trajector. And this rising motion, by prospective an
ticipation, already contains wthin itself the future decline to which
the subject can acquiesce. A kinetic totality is evoked by a reersal of
the subject/object , free/determined polarities wthin a purely spatial
and representational schema.
In "Archaischer Torso Apollos" (1 : 313) the reversal is ocular.
The obserer is, in its turn, being obsered by the fragmentar statue
which has been transformed into a single, large ee: "denn da ist
keine Stelle, / die dich nicht sieht . " The reversal is possible only
because the sculpture is broken and fragmentar; if the statue had
actually represented the eye of Apollo, the chiasmus could not have
come about . The absent eye allows for an imaginar vsion to come
into bing, and it maes the eyeless sculpture into an Agus ee
capable of engendering, by itself, all the dimensions of space. We
always re-encounter versions of the same negativ moment : the hol
low of the volin, the irreality of the mirrored image, the darkness of
a sundial at night, the falling ball, the missing ee. The absences create
the space and the play needed for the reversals and fnally lead to a
totalization which they seemed, at frst , to make impossible. The
broken statue becomes more complete than the intact one, decadent
could not remain in objects / e . that gided into you") (that is to say, in the ball,
objct of the apostrophe) , is a great deal more complex i n the poem than in the
schematic summar we give here for reasons of economy. A detailed reading of"Dr
Bll" would show that we are indeed dealing wth such a reveral of the subjective
"content . "
2. "[the ball] . . . bws dow / lingers and suddenl, from abve, / points
the player to a new place / ordering place and player as in a fgured dnce . . .
TROPES ( RI LKE) 45
Brge richer than the prosprous realit of the past , the falling ball
"happier" than the rising one, the nocturnal dial a more complete
timepiece than the sundial at midday, etc.
The unifng principle of the Ne Po resides in the homol
og of their rhetorcal strcture. Even when the evoke entities
which, unlike a ball , a fountain, a cat , or a gazele, are no longer
relatively ordinar but transcendental or even divine, the strctue
remains the same. As a matter offact the predicates of ordinarness
and transcendence are themselves one ofthe most string reversals.
Rilke desGrbes the rose window of the Chatres cathedral both as the
reabsorption of all existence into the oneness of G and as the eye of
a cat ("Die Fensterose," 1:257) . The shock of this juxtaposition does
not actually deepen our knowledge and understanding of realit and
of Gy but it seduces the mind by the surprise of its precision. I t
captures and fascinates attention by the same skill that allows for the
virtuosity ofits play. It would therefore b a mistake to follow till the
end those commentators who read the NePom a a messianic
text ,27 seeing them as a hierarchized network of sbolic relation
ships that ascend towards the parousia of an omnipresent being. The
numerous successful poems that appear in the volume are primarily
successes of language and of rhetoric. This is hardly surprsing, since
it has been clear from the start that the Rilkean totalizations are the
outcome of poetic skills directed towards the rhetorical potentialities
of the sigifer.
This reversal ofthe traditional prorit, which located the depth
of meaning in a referent conceived as an oect or a consciousness of
which the langage is a more or less faithful refection, asserts itself
in Rilke's poetr by disguising itself at once into its opposite. Ver few
of the Ne Poem openly refer to langage (as was the case wth the
"Eigng" pom of Th Bk ofIme) , but the priorit of1 over
lo
g
os is always apparent in their strcture. Rilke's vocabular retains
this shif in the emphasis and in the authorit of the fgural struc-
27. Such as, for example, the most attentive interpreter of Ne Pom, Hans
Brendt, in Rinr Ma Rlk Neu Ght: Veuh enr Dutun ( Bonn, 1957) .
The recent study b Brigitte Bradley, Rinr Ma Rlk Neu Gdht: Ihr zyk
lih Ge ge (Brn, 196) , is not messianic but dos not attempt an interpretation
of the bok as a whole.
46
RHETORI C
tures when he uses, wth considerable precision, the term "fg"
(Fig) to distingish his rhetorical strateg from that of classical
metaphors.28 B sugesting the potential identifcation of tenor and
vehicle, the tradtional metaphor stresses the possible recuperation
of a stable meaning or set of meaning. It allows one to see langage
as a means towards a recovered presence that transcends langage
itself. But what RIe calls fge is, on the thematic level , anyhing
but a recuperation. The allegor of fguration in a text such as "Or
pheus. Eudice. Henes" ( 1 :298) contributes to the understanding
of this distinction.
The poem explicitly describes the poetic vocation by means of a
thematized version of chiasmic reversal , source of Re's afnity
with the myh of Orpheus. The theme appears twice in the text and
alows one to distinguish the "right" reversal at the end from the
"wrong" reversal described in section III:
Und seine Sinne waren wie entzweit :
indes der Blick ihm we ein Hund vorauslief,
umehrte, kam und immer weder weit
und wartend an der nachsten Wendung stand,
blieb sein Gh6r we ein Grch zurck. 29
This moe of reversl , to which Orpheus wll fally succumb, id
cates the impatience and the desire for a possession within presence.
The absence of being-the death of Eurdice-is the origin of a
desire which expresses itself in the elegiac tonalit of the complaint .
In a passage that prefgres the central theme of the Tenth Duino
Eleg, the complaint is defned as a language capable of creating and
flling an entire potic universe:
Die So-Gliebte, dass aus einer Leier
mehr Klage km als je aus Klagefrauen;
2. On the concept of fge in Rle, the study by B Alleman, Zit un
Fr bm spitn mik, remains indispnsable. (Se note 5 above.)
. 29. "Ad his senses were as doubled: / because his sight, like a dog, ran ahead
of him, / ted around, came back to him and stod / waiting for him at the next
roadbnd,-/ his hearing tarried as if it were an odor. "
TROPES ( RI LKE)
dass eine Wel t aus Klage ward, i n der
alles noch einmal d war: e
d
[ 1 :300]
47
However, since it stems from a desire for presence, the com
plaint is almost inevtably transformed into the impatience of a
desire. It tends to consider the fctional world it engenders as an
absent reality, and it tries to repossess what it lack as if it were an
exterior entit. The confusion can only lead to the loss of language
which, in the symbolism of the poem, coresponds to Orpheus's
increased inabilit to perceive sounds to the point of forgetting the
existence of his lye. To the extent that metaphor can be thougt of
as a language of desire and as a means to recover what is absent , it is
essentially anti-poetic. The genuine reversal takes place at the end of
the poem, when Hermes turns away from the ascending movement
that leads Orpheus back to the world of the living and instead fol
lows Ewydice into a world of privation and nonbeing. On the level of
potic language, this renunciation corresponds to the loss of a pri
mac of meaning located within the rferent and it allows for the
new rhetoric of Rlk's "fge. " RIe also calls this loss of referen
tialit by the ambivalent term of "inwardness" (innn entstehen,
Weltinnnraum, etc. ) , which then does not designate the self
presence of a consciousness but the inevtable absence of a reliable
referent. It designates the impossibilit for the language of poetr to
appropriate anyhing, be it as consciousness, as object , or as a sy
thesis of both.
From the perspective of the language of fguration, this loss of
substance appears as a liberation. It trigers the play of rhetorical
reversals and allows them the freedom of their play wthout being
hampered by the referential constraints of meaning: RiIe can assert ,
for instance, that the refection is more real than realit, or that the
sundial records the hours of the night , because his statement now
exists only in and by itself The same freedom also allows him to
prefgure a new totalit in which the fges will perfectly comple
ment each other, since the totality does not have to take into account
30. "Bloved, so-beloved, that from one lye / Came more woe than ever came
from wailing women / and thus arose a world of wo in which / all thing once more
were present . . o
48
RHETORI C
any empirical or transcendental veracit that might confict with its
principle of constitution. And it also allows for a perfect aticulation
of the semantic wth the rhetorical and phonic function of langage,
thus preserng the initial sound-centered manner as a principle of
poetic compsition. From NePom on, RIe's poetr wll live of
the euphoria of this recovered freedom. A constant refnement ,
which goes far enough t o recover a semblance of simplicit, will
reduce the diversit of fgration that appeas in Ne Pom to a
small number of elective fgres that are particularly productive in
their internal reversals as well as capable of combining with each
other in at t imes dazzling constellations. But the poetr will be able
to achieve this master only at the cost of a subterfuge to which it
fnds itself necessarily condemned.
For this "librating theor of the Signifer"31 also implies a com
plete dng up of thematic possibilities. In order to be a pue poetr
of what RIke calls "fgs," it should start on the far side of the
renunciation which opens up its access to this new freedom. But
could any poetr, including RIke's, lay claim to the purit of such a
semantic askesis? Some of RIe's allegorzing poems, such as "Or
pheus. Eurdice. Hermes" or the Tenth Duino Eleg, programmati
cally thematize the renunciation in a narrative mode, by telling the
stor of this renunciation. In a more lyrical vein, RIke attempted
poems that tend towards the impersonality and the detachment that
should characterize a poetics of pure "fgure. " In those poems, an
emblematic object is revealed to be a fgure without the need of any
discourse, b the ver strcte of its constitution. Such poms appar
in his work from Ne Poem on and will recur till the end, including
some of the poems written in the French language. These poems are
by necessi t brief and enigmatic, often consisting of one singe sen
tence. One might well consider them to be RIke's most advanced
potic achievement . It is through them that he is related to poets
such as Trak or Celano The fgue stripped of any seduction besides
that of its rhetorical elasticit can form, together with other fgures,
constellations of fgures that are inaccessible to meaning and to the
senses, located far beyond any concern for life or for death in the
hollow
space of an unreal sky.
But next to these short and necessarily enigatic tests, RIe has
31 . The expression comes from Rland Brthe and appears on the cover of
S/Z ( Paris, 1970) .
TROPES ( RI LKE) 49
also produced work of a wider, at times monumenta, scope that
are more accessible to understanding. The example of predecessors
such as Holderlin or Baudelaire may wel have guided him in this
direction. The trend is apparent in some of the longer NePoem
and it culminates in the Duir Elg, the work that, more than any
other, has determined the reading of RIe as a messianic poet . For
rather than being themselves poetic fgres, the El state a
genuine existential philosophy of fguration, presented as if it were a
coherent principle of inner behavior, with rules and precepts that
could be set up as exemplar. In principle, the imperative tone of the
Elg is totally incompatible wth the ver notion of pure fg,
which implies the complete renunciation of any normative pathos or
ethical coercion. But there representational and subjective elements
openly play a determining part . Athough the advocate a conception
oflanguage that excludes all subjective or intersubjective dimensions ,
the Duino El constantly appeal to the reader's emotion and par
ticipation.
This paradox is not due to bad faith or to deliberate deception
on the part of RIe; it is inherent in the ambivalence of poetic
language. The primac of the signifer, on which RiIke' s phonocentric
poetics of chiasmus is predicated, is not just one proper of language
among others that would have remained unnoticed during several
centuries until particularly perceptive poets such as Mallarme or
Rilke would have rediscovered it . The notion of a language entirely
freed of referential constraints is properly inconceivable. Any utter
ance can always be read as semantically motivated, and from the
moment understanding is involved the positing of a subject or an
object is unavoidable. In RiIe's major work, the Duir El and,
to a lesser extent, the Snn to Orhu, the relapse from a rhetoric
of fgation into a rhetoric of signifcation ocurs in a way that the
structural description of the Ne Poem made predictable.
Chiasmus, the ground-fgure of the NePoem, can only come
into being as the result of a void, of a lack that allows for the rotating
motion of the polarities. A long as it is confned to objects, this
structural necessit may seem harmless enough: the declining mo
tion of a fountain or of a ball , the reflection of a mirror or the
opening of a window casement have, in themselves, nothing of
pathos about them. But RiIe's fguration must also involve sub
ject/object polarities, precisely because it has to put in question the
. . ih mpelling polarit. This implies
I nstl tut f Or Ro m2ni :.- nl c Phi / ol ogla
I r: v. : " 1 . _80J.bQ(
\=
* ^" 7 ^ + - ~+~
50
RHETORI C
the necessity of choosing as fgures not only things but personal
destinies or subjective expriences as well , with the avowed pupose
of converting them into impersonal over-things, but without being
able (or wanting) to prevent that the subjective moment frst func
tion on the level of meaning. However, these experiences, lie the
fgral objects, must contain a void or a lack if the are to be con
verted into fgres. It follows that only negative experiences can be
petically useful . Hence the prevalence of a thematics of negative
expriences that will proliferate in Rlke's poetr: the insatiability of
desire, the powerlessness of love, death of the unfulflled or the
innocent , the fragility of the earth, the alienation of consciousness
all these themes ft Rlke's rhetoric so well , not bcause the are the
expression of his ow lived experience (whether they are or not is
irrelevant) but bcause their structure allows for the unfolding of his
patterns of fgration. And j ust as the knetic totalization had to
encompass rising and falling motions into one singe trope, or j ust as
the refective totalization must include both sides of the mirror, so
the totalization of subjective experience must lead to a positive asser
tion that only chiasmus can reveal . The reversal of a negativity into a
promise, the ambivalent thematic strateg of the Duirw El, al
lows for a linguistic play that is analogous to that in the most discreet
of the Ne Poem. They call, however, for a vr diferent tone,
whose pathos, feror, and exaltation make one forget the formal and
fctional nature of the unity the celebrate. It is inevitable that the
Elg are being read as messianic poms: all their thematic asser
tions confrm this claim, and it is borne out by the virtuosity of the
fgration.
32
Yet the promise asserted by these texts is gounded in a
play of language that can only come about because the poet has
renounced any claim to extra-textual authority. In conformity with
a paradox that is inherent in all literatue, the poetr gains a
maximum of convincing power at the ver moment that it abicates
any claim to truth. The Elg and the Snn have been the main
32. Jacob Steiner, the most exhaustive i nterpreter of the El (Rlk
Duinr El [Brn, 1962] ) , constantly warns against the tendenc to read too
literally many of the passages which allow for an interpretation of the El H a
t of secular slvation (see Steiner, pp. 160, 210, among others) . The fact remains,
for Steiner, that the convergence btween the ptic achievement and the exstential
depth is never in question. The fnal afmation i seen in all its difcult, to the
point of making its formulation impossible, but this only strengthens its afmative
pwer.
TROPES ( RI LKE) 51
source of evidence in trng to prove the adequation of RIe's
rhetoric to the truth of his afmations, yet his notion of fgural
language eliminates all truth-claims from his discourse.
It would be a mistake to believe that a de mystifing reading of
Rlke could reduce this contradiction to a passing aberration. The
messianic reading ofRIe is an integral part of a work that could not
exist wthout it. The full complexit of this poetr can only appear in
the juxtaposition of two readings in which the frst forgets and the
second acknowledges the linguistic structure that makes it come into
being. The question remains whether RIe himself considered his
work under this double perspective or whether he followed the
example of his commentators in systematically stressing the former
at the expense of the latter.
Some of the particularly enigmatic poems from Re's last perod
cannot easily be reconciled with the positive tonality that is generally
associated, even at this same late date, with the theme of the fgure.
This is the case of the followng poem from the Snnt to Orphu, a
text that has proven to be ver resistant to interpretation:
Sieh den Himmel . Heisst kein Sternbild "Riter"?
Denn dies ist uns seltsam eingepragt :
dieser Stolz aus Erde. Und ein Zweiter,
der ihn treibt und halt und den er tragt .
1st nicht so, gejagt und dann gebandigt,
diese sehnige Natur des Seins?
Weg und Wendung. Doh ein Druck verstandigt .
Neue Weite. Und die zwei sind eins.
Aber sin sie's? Oder meinen beide
nicht den Weg, den sie zusammen tun?
Namenlos schon trennt sie Tisch und Weide.
Auch die sternische Verbindung trgt .
Doch uns freue eine Weile nun
der Figr zu gauben. Das genigt .
[Snnt, 1 :493]33
33. "Behold the sky. Is there no constellation called 'Horseman'? I For we have
been taught , singularly, to expect this: I this pride of earth, and his companion I
who drives and holds him, and whom he carres. II Is he not , thus spured and then
52 RHETORI C
Although i t does not have the somewhat doctrinal tone of some
texts with a similar theme, the poem is important for an understand
ing of Rlke's poetics , since it deals with the recurrent and central
fge of the constellation. The constellation sigifes the most inclu
sive form of totalization, the recupration of a language that would
b capable of naming the remaining presence of being beyond death
and beyond time.
The recovered unity comes into bing in the play of polarities in
the two quatrains, in which we pass from a movement of constraint
and opposition to the condition of acquiescence which we have fre
quently encountered in our readings. The horseman and his steed are
frst shown in a relationship of duality in which their wills combat
each other. The horse' s pride rebels against the will of the rider,
despite the fact that he is entirely at the mercy of the natural and
earthlike power that carries him.
34
The track [ Weg] , the path freely
chosen by the animal , and the turn [ Wend] , which desigates the
will to direct it in a direction of the rider's choice, are at frst in
conflict with each other. This way of being in the world is character
istic of man, a creature that exsts in constant opposition to the spirit
of the earth that inhabits plants, animals, and innocent bings. The
theme of this alienation, of a human destiny constantly opposed to
the natural motion of things, runs through the entire work:
Dieses heisst Schicksal : gegeniber sein
und nicht als das und immer gegeniber.
[Eighth Eleg, 1 :471]35
reined in, I like the nerlike nature of Bing? I Track and turn. But a pressure bring
them together. I New expanse-and the two are one. II But are the trly? Or is the
track they I travel together nt the meaning of thei way? I Table and pasture part
them more than names. II Star-patterns may deceive I but it pleases us, for a while, I
to blieve in the fgre. That i enough. "
34. The sytax of the passge is difcut and has made the task of the com
mentators and of the translators an uncertain one. In agreement with Jacob Steiner,
we read "Dieser Stolz aus Erde" as meaning "this pride made of earh," and as
desigating the horse. "Ein Zweiter" then refers to the rider. The literal meaning is
"and a second one (the rider) who spurs and reins in ( the horse) that caries him."
The rest of the interpretation difers from that of Steiner and of Morchen on several
points.
35. "This is called destiny: to b oppsite thing I and not hing els and always
o
ppsite. "
TROPES ( RI LKE) 53
Such a mode of exsting is said to conform to the "nerelike"
Csehnig) , tough, and resistant natue of being, which lines 5 and 6
put into question:
1st nicht so, gejagt und dann gebandigt ,
diese sehnige Natur des Seins?
The answer to this question has to b negative, for Rlke never
conceives of his relationship to the world, nor especially of his rela
tionship, as pot , to words, as a dialectical one. His entire strateg is
instead to let the poetic meaning b carred by the rhetorical and the
phonic dimensions of language: the seductions of the sytax and of
the fgration have to mae even the most extreme paradoxes appear
natural. The "track" of the meaning and the "turn" ofthe tropes have
to be reconciled by and within the fgure. The poem isolates and
retains this moment in the paradox of a benefcent constraint : "doch
ein Druck verstandigt . " The phrase seizes the instant where the con
trar wills are reconciled by a virtuosity that acquires the gracefl
ease of an apparent freedom. The contrar wills cross over and
change place, following the same shift in point of view that made the
player acquiesce to the descending motion of the ball . The freedom
at once opens up a new free space and reveals a new totalit: "Neue
Weite. Und die zwei sind eins. " This new totalit prefgures the
passage from the earthlike couple to the fgural constellation of "The
Horseman. "
Once this point has been reached, most of RIe's poems would
stop and celebrate the new relationship to the world which the
fgation has revealed. This is what happens, for instance, in the
poem from the Snnt t Orhu that immediately follows upon
this one:
Heil dem Gist, der uns verbinden mag;
denn wir leben wahrhaf in Figen.
[ 1 :494] 36
The second part of the Horseman sonnet , however, puts in ques
tion all that has been achieved and reduce the unifed totalit to a
36. "Hail to the spirit that may bring us together / for we live truly among
fgures."
5
RHETORI C
mere illuion of the senses, as trivial and deceiving as the optical
illusion which makes us prceiv the chaotic dissemination of the
stars in space as if they were genuine fgues, genuine designs traced
upon the background of the skes. "Auch die stemische Verbindung
tr gt": the imaginar lines that make up actual as well as fctional
constellations (the fgural constellations of Rlke's poems) are mere
decei t, false surfaces . The fnal afrmation, "Das genigt ," especially
when compared to the ferent promises that appear in other poems,
seems almost derisive. Far from bing, as is the case in the opening
lines of the Ninth Eleg, a celebration of the moment, it sounds like a
disenchanted concession. One can understand the disappointment of
one of Rlke's ferent commentators , a true believer in his poetic
annunciation: "What are we to think ofthis odd complacenc, which
suddenly seems to satisf itself, and 'for a moment,' wth prosional
and deceptive hopes?
,
'37
Wat is most important in this unexpected thematic turn is that
it comes about at the precise instant when the text states its aware
ness of its linguistic structure and designates the event it descrbes as
an event of language. Not only is the horseman referred to by the
metalinguistic term "fgre," but the unity is stated in terms that are
borowed from the semantic function of langage: "Oder minn
beide I nicht den Weg, den sie zusammen tun?" The lines are
difcult to interpret , but the emphasis on signifcation and on mean
ing is undeniable.
The failure of fgration thus appears as the undoing of the
unity it claimed to establish between the semantic function and the
formal structure of langage. Again, one of the Ne Poem may be
the most economical way to make the fgre of the "road," which
horseman and steed are said to travel together, more comprehensi
ble. The poem entitled "Dr Ball" descrbs the road, the trajector of
the ball ; one could say that it sigi this trajector, that the trajec
tor is the meaning of the poem as its referent . Moreover, the formal,
syntactical structure of the single sentence that maks up the text
exactly mimics the meaning: the sentence climbs and falls, slows
dow, hesitates, and speeds up again in a manner that parallels at
all points the signifed motion. The manner of enunciation corre
sponds exactly wth what is being said. In other poems, the same
convergence wll be achieved by way of phonic rather than sytacti-
37. Morchen, p. 122.
TROPES ( RI LKE)
55
cal elements. The logcal meaning and the li indeed travel along
the same road.
But can it be asserted that this parallelism signifes, in the full
meaning of the term, the unity that it constitutes? Is it not rather a
play of language, an illusion as arbi trar as the shap of the constel
lations which share a common plane only as the result of an optical
appearance? The Horseman sonnet confrms that Rlke ke this to
be the case: the fgre's truth turns out to be a lie at the ver moment
when it asserts itselfin the plenitude of its promise. The sonnet is not
the only instance of such a retreat. In a late text entitled "Gng"
(2: 186) Rlke attempts the ultimate reversal , not just the visual rever
sal that takes place in "Achaischer Torso Apollos," but the reversal
within the phonic dimension, wthin the ear, itself: "Kang, / der, wie
ein tieferes Ohr, / uns, scheinbar Horende, hart e e e e "31 Yet , in this
poem, the accumulation of the most extreme paradoxes and of ulti
mate reversals does not lead to the expected totality, but ends instead
in the ignominy of a fall which has nothing in common with the
happy descent of the ball. It suggests instead the denunciation of
the ultimate fgure, the phonocentric Ear-god on which Rilke, from
the start , has wagered the outcome of his entire poetic success , as
error and betrayal :
Wanderers Sturz, in den Weg,
unser, an Alles, Verat . . . : Gong!39
Among Rilke's French poems which, by their use of a foreign
language, correspond to the renunciation of the euphonic seductions
oflanguage, one fnds the same defnition of the fgure as the conver
sion of representational and visual into purely auditive rhetoric:
ing.
II faut fermer les yeu et renoncer i la bouche,
rester muet , aveugle, ebloui :
L'espace tout ebranle, qui nous touche
ne veut de notre etre que 1'0uie. 40
38. "Sund, / which, as a deeper ear, / hears us, who appar to be hear
1
39. " Fall of the wanderer, on the roadside / Our, ofevelhing, btrayal . . . :
Gng!"
40. "We must close our eyes and renounce our mouths, / remain mute, blind,
dazzled: / Vibrat i ng space, as it reaches us / demands from our being only the ear. "
56
RHETORI C
At the moment of its fulfllment , this fgure announces itself by its
real name:
Masque? Non. Tu es plus plein,
mensonge, tu as des yeux sonoresY
More still than the thematic statement, which can alway be
interreted as a recuperation of the posited theme beyond its most
absolute negation, the shift to French indicates not only the kowl
edge but the advent of the disruption. The promise contained in
Rilke's poetr, which the commentators, in the eagerness of their
belief, have described in all its severe complexit, is thus placed, by
Rilke himself, within the dissolving perspective of the lie. RiIe can
only be understood if one realizes the urgenc of this promise to
gether with the equally urgent , and equally poetic, need of retracting
it at the ver instant he seems to be on the point of ofering it to us.
41. "Mask? No. You are fuller / you lie, you have sonorous eyes. "
3 Reading
(Put)
GEORGES POULET HAS TAUGHT US TO CONS I DER, I N
A l rehrch du tmps perdu, the jutaposition of diferent temporal
layers rather than the umediated experience of an identit, gven or
recovered by an act of consciousness (involunta memor, proleptic
projection, etc. ) . t The specifcit of Proust's novel would instead be
grounded in the play beteen a prospective and a retrospective
movement. This alternating motion resembles that of reading, or
rather that of the re-reading which the intricac of ever sentence as
well as of the narrative network as a whole constantly forces upon
us. Moreover, as Poulet descrbes it, the moment that mark the
passage from "life" to wrting corresponds to an act of reading that
seprates from the undiferentiated mass of facts and events, the
distinctive elements susceptible of enterng into the composition of a
text . This occurs by means of a process of elision, transformation, and
accentuation that bears a close resemblance to the practice of critical
understanding. The intimate relationship between reading and criti
cism has become a commonplace of contemporar literar study.
What does A l rehrch du temps perdu tell us about reading? I
approach the question in the most literal and, in fact , naive way
possible b reading a passage that shows us Marcel engaged in the
act of reading a novel. This procedure in fact begs the question, for
we cannot a prri be certain to gain access to whatever Proust may
have to say about reading by way of such a reading of a scene of
reading. The question is precisely whether a literar text is about that
which it describes, represents, or states. If, even at the infnite dis
tance of an ideal reading, the meaning ra is destined to coincide
with the meaning statd, then there would in fact be no real prob
lem. All that would be lef to do would be to allow oneself to be
brought nearer to this ideal perfection by takng Marcel for our
1 . Se Poulet's essy "Poust prospectif" in Meure d ['ntnt (Pars, 19)
and als his L'epae Putin ( Paris, 193) .
57
5 RHETORI C
moel . But i f reading i s truly problematic, i f a nonconvergence be
tween the stated meaning and its understanding may b suspected,
then the sections in the novel that literally represent reading are not
to be privileged. We may well have to lok elsewhere, in Marcel's
erotic, political , medical , or worldly experiences, to discover the dis
tinctive structures of reading, or we may have to go further afeld still
and use a principle of selection that is no longer thematic. This
circular difcult should not , however, prevent us from questioning
the passage on actual readin

.
i
f

p
nly to fnd out whether or not it
does make paradigmatic ims for itself. The uncertaint as to
whether this is indeed the case creates a mod of distrust which, as
the later stor of Marcel's relationship wth Abertine makes clear,
produces rather than paralyzes interpretative discourse. Rading has
to begin in this unstable commixture of literalism and suspicion./f j J)'l..
The main text on reading occurs early in the novel , in the frst
volume ofDu Cot d ch Swann ( 1 :82-8) . 2 It stands out as distinctly
marked in the narrative of "Comb ray" where it follows immediately
upon the young Marcel's vsit to his uncle, the frst explicit example of
his ritualistic initiation to the ambivalences of goo and evi l. The
scene is set within a thematic of closeted and hidden spaces, the
"temple of Venus" of Framoise's bower (p. 72) , the "dark and fresh"
smelling closet in which Uncle Adolphe retires (p. 7) which will
engender a chain of assoiations that wll articulate the entire middle
part of the book, 3 the "dark freshness" of the rom in which Marcel
will hide in order to read (p. 83 , 1. 28) , the "little sentr-box" where
he fnds refuge when his grandmother orders him to go outside
(p. 83, 1 . 4) . The symblic signifcance of this stting is sumarzed
in the interiorized image of the mind as a "cradle at the bottom of
which I remained sheltered, even in order to obsere what was
happening outside" ( p. 8, 11 , 4-5) . The frst section of the passage
(p. 80, 1. 18 to p. 82, 1. 41) does not deal with reading; it is three
pages later when Marcel will climb to his room with a book ( p. 83 ,
1 . 5) , and only when he has been sent into the garden (p. 83, 1 . 41) will
the principal and ver systematically structured discourse on reading
2. Quotations are from Marcel Pout , A l rehrch du tmps pru ( Pars:
Bibliotheque de la Pleiade, 1954) , edited by Pierre Clrac and Ade Fere, in 3
volumes. Pages 8-8 of Volume 1 have ben numbered line b line.
3. By way of the "fresh and closted smell" of the public lavator ( 1 :492) ,
which is also the place where the extended narative ofthe grandmother's death will
bgin.
READI NG ( PROUST)
59
be allowed to develop (p. 8, 1 . 3 to p. 8, 1. 16) . But this preliminar
section is solidly lined to the main body of the passage by a tran
sitional scene centered on the characters ofFrancoise and the ktchen
maid ( p. 82, 1. 18) who was the main figure in the frst section:
"Wile the ktchen maid-unwittingly making Francoise's superior
it shine at its brightest , just as Error, by contrast , makes the
triumph of Truth more dazzling-ered cofee which, in my
mother's judgment , was mere hot water and then caried to our
rooms hot water that was barely tepid, I had stretched out on my
bed, with a book e e . "4 The allegorical pair of Truth and Error
crowns a passage that will be particularly rich in rotating polarities.
But here, in this context of comedy, the chain of substitutions in no
way preseres the integrit ofthe point of origin: the tepid liquid is a
lowly version of genuine hot water, itself a degraded substitute for
cofee. The ktchen maid is only a pale refection of Francoise; in
substituting for truth, error degades and Qutweas it , causing a se
quence oflapses that threatens to contaminate the entire section. All
the later polarities wll have to be on the defensive when placed
under the aegis of the initial antithesis between ruth and error.
Thus reading is staged, from the beginning of the text , as a
defensive motion in a dramatic contest of threats and defenses: it is
an inner, sheltered place (bower, closet , room, cradle) that has to
protect itself a

inst the invasion of an outside world, but that never


theless has to Dgtw from this world some of its properties. The
inside room "tremblingly shelters . . . its transparent and fragle
coolness from the afternoon sun" (p. 83, 1. 6) . The inner world is
unambiguously valorized as preferable to the outside, and a consis
tent series of attractive attributes are associated wth the well-being
of the enclosed space: coolns, the most desirable of qualities in this
novel of the "solar myth" in which the baometer s ofen indicates fne
weather, itself linked to the restorative drkns of shaded light
( Marcel being nevr so happy as when he dwells in the shade of the
vegetal world) , and fnally trnuilit, without which no time would
be available for contemplation. But Marcel cannot rest satisfed with
these positive aspects of a sedentar solitude. The truly seductive
force of the passage is revealed only when the confnement to the
obscure, private existence of inward retreat turns out to be a highly
4. A god two pages earlier (1:8, L 21) Fram;oise has said: "I'll let my ktchen
maid sere cofee and bring up hot water . . q
.
' t v
h" ,
\'
60
RHETORI C
efective strateg for the retrieval of all that seemed t o have been
sacriced. The text asserts the possibility of recuperating, by an act of
reading, all that the inner contemplation had discarded, the oppo
sites of all the virtues necessar to its well-being: the warth of the
sun, its ligt, and even the ativit that the restful immobilit seemed
to have defnitively eliminated. Miraculously enriched by its anti
thetical properties, the "dark coolness" of the room thus acquires the
light without which no reading would be possible, "the unmediated,
actual , and persistent presence" of the summer warmth and fnally
even ". . . the shock and the animation of a food of activty [un
trent d'ativit] . " The narator is able to assert , without seeming to
be preposterous, that by stayng and reading in his room, Marcel's
imagination fnds access to "the tptal spectacle of Summer," includ
ing the attractions of direct physical action, and that he possesses it
much more efectively than if he had been actually present In an
outside world that he then could only have kow by bits and pieces .
Two apparently incompatible chains of connotations have thus
been set up: one, engendered by the idea of "inside" space and
governed by "imagination," possesses the qualities of coolness, tran
quilit, darkness as well as totalit, whereas the other, linked to the
"outside" and dependent on the "senses," is marked by the opposite
qualities of warmth, activt, light , and fragmentation. These in
itially static polarities are put in circulation by means of a more or
less hidden system of relays which allows the properties to enter into
substitutions, exchanges, and crossings that appear to reconcile the
incompatibilities of the inner with the outer would. 5 Proust can
5. Similar fgures, ofen polarized around systems of light/dark and in
side/outside, ar so frequent that they could be sid to make up the entie novel .
The occu from the fst sentence, which has to do wth light and dark, truth and
error, wake and sleep, prception and dream, and which tus on a literaliztion of
the fundamental epistemologcal metaphor of understanding as seeing. One of the
most inter
esting examples, also involvng Giotto, ocurs in the later part ofthe novel ,
in a pas
sge from Lfgtiv, durng Macel's visit to Venice, in the compny of his
mother, afer Albertine's death (3:6) . Grad Gnette ("Metonymie chez Proust ,"
now in Fgr III [Paris, 197], p. 4) quotes the passge H an example of diegetic
metaphor, metaphors in which the selection of the vehicle is dictated by the prox
imit of a detail that happns to b present in the narrative context. The blue col or of
the backgrounds in the Giotto frscoes at the Arena of Padua are said to b "so blue
that it looks as if the radiant dayligt had crossed the threshold in the company of
the viitor, and would have housed for a moment its pure sk in the coolness of the
shade, a pure sk hardly darkened by being rid of[debarasse] the golden sunlight , as
READI NG ( PROUST) 61
afect such confdence i the persuasive power of his metaphors that
he pushes stlistic defance to the point of stating the assumed syn
thesis of light and dark in the. i!pntrovertible language of numerical
ratio: "The dark cool of my wa room was to the full sunlight of
the street what the shadow ist the sunray, that is to say equaly
luminous . . . " (p. 83, 1 . 28) . In a logic dominated by truth and error
the equation is absurd, since it is the diference of luminosit that
distinguishes between shadow and light: "that is to say ("c'est a
in these brief moments of rspite that interrpt the most beautiful dys when,
without having seen a single cloud, the sun having turned its eye elsewhere [Ie soleil
ayant toure ailleurs son rgr] for a moment, the blue ofthe sky sofens and turns
darker. " The comparison of the two blues (Giotto's bckground and the sky) stems
indeed from the proxmit of the prous narative setting in the phrae "afer
havig traversed the grden of the Arena under the full sunlight" ( 3:6) and can
thus legtimately b called a diegetic metaphor. But it is clear that more is at stake in
the passge. The initial sitution is ver similar to that ofthe section we are dealing
with, since the psitiv valoriztion of coolnes and shelter (as maked, for instance,
b the negtive connotations of the word "debarasse" which characterizes the full
light as undesirable) indicates that the metaphor attempts a reconciliation of such
incompatible polarities as hot/cold, inside/outside, light/dark, as well as nature/art .
For the light of art, which is devoid of natural warmth and therefore potentially
devoid of life, to b lie natur, it must be able to borrw, by analog, the attribute of
wrth from the sun without losing its desirable colness. Natural light has to cros
the threshold of its spcular representation; this illusion is convincing enoug since
at least some natual light, however shaded, has to pnetrate into the building for
the frescos to b visible. The burden ofthe passge is therefore not so much to injct
warmth into art as to inject coless into natue; otherse the symmetr of the
totalizing chiasmus could not come about . Hence the necessit for an anlogcal
description in which the heat of the sunlight would not be incompatible wth a
degee of coolness. Wat maes the passge remarkble and takes it well beond
Gnette's moel ofa reconciled sstem of metaphor and metonymy (of "liaison" and
"marrage") , is that Pust refuses to avil himself ofthe simple natural analog that
immediately comes to mind and gos out of hi way to insist that the cool darkening
of the sun is nt caused b a cloud. The sentence "Ie soleil ayant toume ailleurs sn
regard" thus bcomes pure nonsense from the naturalistic point of ve\' that the
logic of the passage, strctured as a natre/at dialectic, demands. The implication
are far-reaching, not only for Gnette's model of happy totaliztion, but for the
entire notion of tropolog as a closed sstem. Such sstem depend on the necesr
link btwen the existence and the kowledg of entities, on the unbreakble
strength of the tie that unites the sun (as entit) wth the ee (as the kowledge of
the entit) . The sentence "the sun having tured its ee elsewhere" i therefore, frm
a trpological point of vew, the most impossible sentence conceivable. Its absurdit
not only denies the intelligbilit of natural metaphors but of all trops; it is the
fgre of the unreadabilit of fges and therefore no longer, strictly spang, a
fgre.
RHETORI C
die") in the quotation is precisely what cannot be said. Yet the logic
of sensati on and of the imagination easily remains convnced ofthe
accuracy ofthe passage and has not the least di fcult in accepting it
as legitimate. One should ask how a blindness comes into being that
allows for a statement in which truth and falsehood are completely
su
bverted to be accepted as true wthout resistance. There seems to
be no limit to what tropes can get away wth.
Strctures and relays of this kind, in which properties are sub
stituted and exchanged, characterize tropological systems as bing,
at least in part , paradigtic or metaphorical sytems . Not surpris
ingly, therefore, this introuctor passage on reading that was
placed, from the beginning, under the auspices of the epistemologi
cal couple of trth and error, also contains statements claiming the
priorit of metaphor in a binar sytem that opposes metaphor to
metonymy.6 The passage refects on the modalit of the sun's pres
ence in the room: it is frst represented in vsual terms by means of
the metaphor of a "refection of light which . . . succeeded in mak
ins its yellow wings appear [hind the blinds], and remained mo
tionless . . . poised like a butterfly" ; then in aural terms by the
resonance of "blows struck . . . aginst the dust crates" in the
street, and fnally, still in aural terms, by the buzzing of the flies,
generalized into "the chamber music of summer" (p. 83, l. 20) . 7 The
crossing of sensor attributes in synaesthesia is only a special case of
a more general pattern of substitution that all tropes have in com
mon. It is the result of an exchange of properties made possible by a
proximit or an analog so close and intimate that it allows the one
to substitute for the other wthout revealing the diference necessar
ily introduced by the substitution. The relational link between the
two entities involve in the exchange then becomes so strong that it
can be called necessr: there could b no summer without fies, no
fies wthout summer. The "necessar link" that unites fies and
6. The study of this plarit has ben masterfully bgn by Grard Gnette
(se note 5 abve) . Proust himlf at times maks \ of rhetorical terms such H
metaphor, alliteratio, and anacoluthon, but neer uses, to my knowledge, the word
metonyy.
7. For the purpos of this reading "chambr music" is taken at face value. It i
of cour, in this novel, a higly marked term singled out from the fst page ("un
qutuor" [1 : 3]) to the various k epioes involvng the Vinteuil sonata, the septuor,
Morel, etc. In this passage, and for ou spcifc purps, it sufces to stress the
sothing connotations of music that obliterates whatever unpleasnt assiat ion
my b evoked by swar of buzzing fies; the image function as a reconciliat ion of
the classical antinomy of art and nature.
READI NG ( PROUST) 6
summer is natural , genetic, unbreakable; although the fies are only
one minute part of the total event designated by "summer," the
nevertheless partake ofits most specifc and total essence. The sec
doche that substitutes part for whole and whole for part is in fact a
metaphor,8 powerful enough to transform a temporal continguit
into an infnite duration: "Born of the sunny days, resurrected only
upon their return, containing some of their essence, [the buzzing of
the fies] not only reawakens their image in our memor but certifes
their retun, their actual , persistent, unmediated presence. " Com
pared to this compelling coherence, the contingency of a metonymy
based only on the casual encounter of two entities that could ver
well exst in each other's absence would be entirely devoid of poetic
power. "The tune of human music [as opposed to the "natural" fies]
heard perchance during summertime . . . " may be able to stimulate
memor in a mechanical way, but fails to lead to the totalizing
stabilit of metaphorcal processes. If metonymy is distinguished
from metaphor in terms of necessit and contingenc (an interpreta
tion of the term that is not illegitimate) , then metonymy is per
defnition unable to create genuine lins, whereas no one can doubt,
thanks to the butteries, the resonance of the crates, and especially
the "chamber music" of the fies, of the presence of light and of
warmth in the room. On the level of sensation, metaphor can recon-
` In ' ).
cile night and day in a chirscur that is entirely convncing. But the cW" _v
passage plays for higher staes.
For it does not sufce for the sound of the fies to bring the
outside light into the dark room; if it is to achieve totalization, the
inwardness of the sheltered reader must also acquire the power of a
concrete action. The mental process of reading extends the function
of consciousness beond that of mere passive perception; it must
acquire a wder dimension and become an action. 9 The light
8. Classical rhetoric generally classifes synecdoche H metonyy, which leads
to difculties characterstic of all attempts at establishing a taonomy of tropes;
tropes are transforational systems rather than grids. The relationship between part
and whole can b understod metaphorically, as is the case, .for example, in the
organic metaphors dear to Gethe. Synecdohe is one of the borderline fges that
create an ambivalent zone between metaphor and metonyy and that , by i ts spatial
nature, creates the il lusion of a sythesis by totalization.
9. The use of the term "action" (which stems from Poust's text) does not
mean that metaphor is here conceivd as a spch act . "Activtl" has the meaning of
atu eeritu in a classical plarity of mental contemplation. versus physical action.
A reading of Proust in terms of spech-act theor would have to proeed along
diferent l ines.
6
RHETORI C
metaphors are pwerless t o achieve this: i t will take the interention
of an analogical motion stemming from a diferent propert, this
time borrowed not from the warith of the light but from the cool
ness of the water: "The dark coolness of my room . . . matched my
repose which ( than to the adventures narated in my book, which
stirred my tranquilit) supported, like the quiet of a hand held mo
tionless in the middle of a running brook, the shock and the anima
tion of a food of activit ("mon repos . . . supportait, pareil au
repos d'une main immobile au milieu d'une eau courante, Ie choc et
l'animation d'un torrent d'activi te") . The persuasive power of the
passage depends on the play on the verb "supporter" which must be
strong enough to be read not just as "tolerate" but as "support ,"
sugesting that the repose is indeed the foundation, the ground that
makes activit possible. Rpose and action are to merge as intimately
as the "necessar link" that ties the column to its pedestal .
The ethical investment in this seemingly innocent narrative de
scription is in fact considerable enough to match the intrcac of the
rhetorical strateg. For the burden ofthe text, among other things, is
to reassure Marcel about his flight away from the "real" activty of
the outer world. The guilt pleasures of solitude are made legitimate
because the allow for a possession of the world at least as virle and
complete as that of the hero whose adventures he is reading. Against
the moral imperative speaking through the grandmother who "begs
Marcel to go outside," Marcel must justif his refusal to gve up his
reading, together with all the more or less shameful pleasures that go
wth i t. The passage on reading has to

temf the reconciliation


between imagination and action and to resolve the ethical conflict
that exists between them. If it were possible to transform the imag
nar content of the fction into actions performed by the reader, then
the desire would be satisfed without leaving a residue of bad con
science. An ethical issue that is obviously involved in the success of
the m6ta
g
h
or is connected to the central Poustian motive of guilt
and bet
'
afal that governs the narrator's relationship to himself and
to thos united to him by ties of love or afection. Guilt is always
centered on reading and on wri ting, which the novel s ofen evokes
in somber tones. This connection between metaphor and ilt js one
of the recurent themes of autobiographical fction. 5.L. ,)
One should not conclude that the subjective feelings of guilt
motivate the rhetorical strateges as causes determine efects. It is
not more legitimate to say that the ethical interests of the subject
READI NG ( PROUST) 65
determine the invention of fgures than to say that the rhetorcal
potential of language engenders the choice of gilt as theme; no one
can decide whether Proust invented metaphors because he felt gilty
or whether he had to declare himself guilty in order to fnd a use for
his metaphors. Since the only irreducible "intention" of a text is that
of its constitution, the second hypothesis is in fact less unlikely than
the frst. The problem has to be left suspended in its own indecision.
But by sugesting that the narrator, for whatever reason, may have a
vested interest in the success of his metaphors, one stresses their
operational efectiveness and maintains a certain critical vigilance
with regard to the promises that are being made as one passes from
reading to action by means of a mediating set of metaphors.
In this passage, the metaphorical relay occurs by way of the
flowing water: repose supports action "like the quiet of a hand, held
motionless in the middle of a running brook. " In the sunny mood of
the text, the image is convincing enough: nothing could be more
attractive than this feeling of freshness rising from the clear water.
But coolness, it will be remembered, is one of the attributes of the
"inner" world, associated with shelter, bowers, and closed rooms.
The analogical image of the hand is therefore not able to cross over,
by its own power, towards a life of action. The water carres with it
the property of coolness, but this quality, in the binar logic of the
passage, belongs to the imaginar world of reading. To gain access to
action, the trope should capture one of the properties that belongs to
the antithetical chain such as, for example, warmth. The cool repose
of the hand should be made compatible wth the heat of action. This
transfer occurs, still wthin the space of a single sentence, when it is
said that repose supports "un torrent d'activite. " In French, this ex
pression is not-or is no longer-a metaphor but a cliche, a dead or
sleeping metaphor which has lost its literal connotations (in this
case, the connotations associated wth the word "torrent") and has
only kept a proper meaning. 1 0 "Torrent d'activi te" properly sigifes a
10. Thus illustrating the tripatite structure of all metaphors, ofen stressed by
theoreticians of rhetoric, but not clearly embdied in ordinar English language,
which distinguishes only vaguely between literal and "proper" meaning. Wen
Homer calls Achilles a lion, the literal meaning of the fge signifes an animal of a
yellowish brow color, living in Africa, having a mane, etc. The fgural meaning
signifes Achilles and the proper meaning the attribute of courage or strength that
Achilles and the lion have in common and can therefore exchange. In the cliche
"torent d'activite" (as when I sy of a hyperactive Mr. X that he is O torent
66
RHETORI C
lot of activity, the quantity of activty likely t o agitate someone t o the
point of makng him feel hot . The proper meaning converges with
the connotation supplied, on the level of the signifer, by the "torride"
("hot") that one can choose to hear in "torrent. " Heat is therefore
inscribed in the text in a underhand, secretive manner, thus linkng
the two antithetical series in one single chain that permits the e
change of incompatible qualities: if repose can be hot and active
without however losing its distinctive vtue of tranquility, then the
"real" activity can lose its fragmentar and dispersed quality, and
become whole wthout havng to be any less real .
The transfer is made seductive and convncing b a double
faced play on the cliche "torrent d'activite. " The neighborng image
of flowing water ( the hand suspended "in a running brook") re
awaens, so to speak, the dozig metaphor which, i the clche, had
become the mere continguity of two words ( "torrent" and "activite")
sytagmatically joined by repeated usage and no longer by the con
straints of meaning. "Torrent" fnctions in at least a double seman
tic register: in its reawakened literal sense, it relays and "translates"
the property of coolness actuall present in the water that covers the
hand, whereas in its fgal meaning it designates an amplitude of
action sugestive of the contrar quality of heat .
The rhetorical strcture of this part of the sentence ("repose
. . . supported . . . the shock and the animation of a food of activ
it) is therefor not simply metaphorical . It is at least doubly
metonyic: frst because the coupling of two terms, in a cliche, is
not governed by the "necessar lin" of a resemblance (and potential
identity) rooted in a shared property, but dictated by the mere habit
of proxity (of which Prout , elsewhere, has much to syli) , but as
because the reanimation of the numbed fge takes place by means
of a statement ("running brook") which happens to be close to it ,
without however this proximity being determined by a necessity that
would exst on the level of transcendental meaning. To the contrar,
the property stressed by the neigboring passage is precisely not the
d'activite") the literal meaning of torent has ben lost and only the share attrbute
of "muchness" remains. I. A. Rchards's distinction beteen gound, tenor, and
vehicle desigates this sme strcte. It is pr of our argument that such numercal
and geometrcal moels, assuing the spcifcit of each paricular trop, thoug
unavoidable, are in the long rn intenable.
11 . Se, for instance, the lengthy development on "habitude" at the begnning
of the second part of A l'ombre djeunfll enjurs ( 1 :6 i. ) .
READI NG ( PROUST) 67
property that sered in the coinage of the original metaphor, now
degraded and become a cliche: the fgure "torrent d' activih:?' is
based on amplitude and not on coolness. This propert functions in
fact against the quality that the text desires.
The structure is tical of Proust's language throughout the
novel. In a passage that abounds in successful and seductive
metaphors and which, moreover, explicitly asserts the superior
efcac of metaphor over that of metonymy, persuasion is achieved
by a fgural play in which contingent fgures of chance masquerade
deceptively as fgures of necessity. A literal and thematic reading that
taes the value assertions of the text at their word would have to
favor metaphor over metonymy as a means to satisf a desire all the
more tempting since it is paradoxical : the desire for a secluded read
ing that satisfes the ethical demands of action more efectively than
actual deeds. Such a reading is put in question if one takes the
rhetorical structure of the text into account.
The central text on reading (p. 83, 1 . 38 to p. 8, 1 . 16) develops in the
wae of this initial complication. It has all the appearances of a set
piece, so frmly constructed that it constantly attracts attention to its
own system and invites representation by means of synoptic dia
grams. The text follows "from inside to outside the layers simulta
neously jutaposed in [the] consciousness . . . " of the reader (p. 87,
1. 22) . It extends the complexity ofa singe moment in time upon an
as oriented from maimum intimac to the external world. This
construct is not temporal , for it involves no duration. The diachrony
of the passage, as the narrative moves from a center towards a
peripher, is the spatial representation of a diferential but com
plementar articulation within one singe moment. For a novel that
claims to be the narrative extension of one singe moment ofrecollec
tion, the passage undoubtedly has paradigmatic signifcance. The
transposition of the present moment into a consecutive sequence
would correspond to the act of fction-wrting as the narration ofthe
moment. This act would then be coextensive with the act of self
reading by means of which the narator and the witer, now united
in one, full understand their present situation (including all its
negative aspects) by means of the retrospective recapitulation of its
genesis. Nor would it difer from the response available to the
reader of A l rechrch du tmps perdu who, mediated by Proust's
novel , understands the narrative voice as the dispenser of a true
6
RHETORI C
knowledge that also includes him.
Z
The "moment" and the "narra
tion" would be complementar and symmetrical , specular re
fections of each other that could be substituted without distortion.
By an act of memor or of anticipation, the narrative can retrieve the
ful experience of the moment. We are back in the totalizing world of
the metaphor. Narative is the metaphor of the moment , as reading
i the metaphor of writing.
The passge is indeed ordered around a central , unitg
metaphor, the "single and unbending projection of all the forces of
my life" ("meme et infechissable jaillissement de toutes les forces de
ra ve") within which the various levels of reading are said to consti
tute "sections at the diferent levels of an iridescent fountain that
appeared to be motionless" (p. 87, 11 . 18-19) . The fge aims at the
most demanding of reconciliations, that of motion and stasis, a
synthesis that is also at stake in the model of narrative as the dia
chronic version of a single moment. The continuous fow ('aillisse
ment") of the narrative represents an identit that is beyond the
senes and beyond time as something accessible to sigt and sensa
tion and therefore comprehensible and articulated,just as the unique
and timeless1 4 fascination of reading can be divided into consecutive
layers shapd like the concentric rings of a tree trunk. Within a closed
system of part and whole, the complementarit of the vertical jux
taposition and the horzontal succession is frmly established. With
regard to the narrative, the proof of this complementarit will be the
absence of interruptions, the lack of jaged edges which allows for
the characterization of the novel's narrative texture as a play of
fragmentation and reunifcation that can be called "fondu," (Le. ,
smooth [Grard Gnette]) or "soude," ( i . e. , welded [Proust]) . 1 5 The
12. "In trth, each reader is, when he reads, the actual [prpr] reader of
himself (3:
91 1
) .
13. In a famous passage of Th Plu, Wordsworth spa of "The stationar
blast of watenall
s" (V, I. 626) . A more literal and less benevlent version of this
sme waterpout appars in Sm e Gmr: the fountain designed by Hubrt
Rbrt that splashes Mme. d'Arpajon to the gat merriment of the gand duke
Wadimir (2:657) .
14. " . . . the concentration of my reading, like the magc ofa deep sleep . . .
had erased the ringng of the golden church bells on the sky-blue sunace of silence"
( 1 :8, II. 2-5) .
15. For example, in a passge referring to Vinteuil's sptuor: " . . . to entirely
diferent modes of questioning, the one breang up a pure and continuous line into
brief requests, the other welding [sount] stray fragments into one single, sturdy
READI NG ( PROUST) 69
continuit is not ony apparnt in the fuenc of the transitions or in
the numberless symmetres of the composition, but also in the strict
coherence between meaning and strcture. The passage is a persua
sive case in point: to the stated assertion that reading is gounded in
a fm relationship between inside and outside corresponds a text
that is structred in a particuarly rgorous and systematic way. But
if the complementarit were to b an illusion, a ver diferent stor
would ensue, more lie the loss of entropy that occus as one moves
from Franoise's hot cofee to the kitchen maid's tepid shaving
water.
The peruasive value of the passage depends on one's reading of
the foutain as an entit which is both immobile and iridescent. The
iridescence is prefgured a few pages ealier in the description of
consciousness as a "shimmering screen" ("un ecran diapre") (p. 8,
1 . 13) . The miraculous interference of water and light i n the refracted
rainbow of the color spectrum makes its appearance througout the
novel , infallibly associated with the thematics of metaphor as totali
zation. I6 It is the perfect anaogon for the fgure of complementarit,
the diferences that make up the parts absorbed in the unit of the
whole as the colors of the spectrm are absorbed in the original
white light . The solar myh of A l rhrch du tmps perdu would
then be condense in the scarf of Irs, as when the flower metaphors
associated with girls and women are said to "appear at once on their
two sides, lie complementar colors" (p. 86, 1. 20) . The "necessa
lin" between the imagned fge and its sensor qualities mae it
more seductive than the empirical , "real" landscape of Combray.
frame" (3:255) . Grard Gnette ("Metonyie chez Proust," p. 60) mentions a pas
sag from Prust's c9rrespondence (Crpon [Pas, 1970] , 2: 86) which uses
the expression "espce de fondu. "
16. Some examples among many others: Elsti's workhop i compared t o a
"lo of ro -rtal, of which one o the facts, aldy ct and plihe, shie lie
an irdescent miror" ( 1 :835) ; Fram;oie's famous asparagus "reveal in thei nascent
colors of early dawn, in their sugestions of rinbows . . . [their] costly essnce"
( 1 : 121) ; "if I coud have analyzed the prim [of the duchess de Guerante's ees]
. . . the esence of the unow life that appeard in them might hav ben
revaled to me" (2: 53) ; "the art ofVinteuil, lie that of Elstir, reveals [the inefable
character of indivdualit] by expressing into the colors of the spectrm the

intimate
being of the worlds we call individuals " (3:258) ; 'just as the spectrm rpr
snts for us the compsition of light, the harmony of a Wagner or the color of
an Elstir allows us to kow the qualitative essnce of another indivdual's sns
tions . + . " (3: 159).
10
RHETORI C
Unlike this real landsape, the symblic one is "a true part of Nature
itself, worthy of study and meditation" (p. 86, 1 . 34) .
The supriorit of the "symbolic" metaphor over the "literal ,"
prosaic, metonymy is resserted in terms of chance and necessit.
Within the confnes of the fction, the relationship beteen the
fgures is indeed govered by the complementarit of the literal and
the fgral meaning of the metaphor. Yet the passage seems oddy
uable to remain sheltered within this intra-textual closue. The
complementarit is frst assered with reference to the narrator's
relationship to the landscape he inhabits, but it soon extends towards
another binar set of themes, those of "love" and "voyage": ''There
fore, if I always imgined, surrounding the woman I loved, the
landscape I most keenly wished to see at that moment . . . it was
not because a mere association of ideas exsted between them. No, it
is because my dreams of love and of travel were only moments
which I now artifcially disentagle . . .-in the single and unbnd
ing projection of all the forces of my life" (p. 87, 11. 11-21) . But what
is here called "love" and "travel" are not , like the narrator and his
natual setting, two intra-textual moments in a fction, but rather the
irresistible motion that force any text beyond its limits and projects
it towards an exterior referent. The movement coincides with the
need for a meaning. Yet at the begnning of the passage Marcel has
stated the impssibilit for any consciousness to get outside itself,
sugesting this ver idealit, paradoxically enoug, by means of an
analog derived from a physical phenomenon: "Wen I saw some
thing external , my awareness of the fact that I was seeing it re
mained between the object and myelf, borering it as with a thin
spiritual layer that prevented me from touching it drectly; the object
would evaporate, so to speak, before I could come into contact with
it , just as a red-hot body that approaches a wet object is unable to
touch its humidity, since it is always preceded by a zone of vapor"
(p. 8, 11. 5-13) . Three pages further on, it sems that the language of
con
siousness is unable to remain thus ensconced and that, like so
many objects and so many moments in Proust's novel, it has to tun
itself out and become the outer enveloping surace: 1 7 "For if we have
17. The metonymy b which the covered-up entit becomes its own cover
[evlppe becomg evlppant] is much in evdence in the concluding section of
this passge, where "the aferons have gadully surounded and enclosed" the
hour: the spatial container bcomes the temporally contained, and vce vers. The
famous passage on the "carafes de l Vivonne" ( 1 : 16) i the lcu cliu of thi
READI NG ( PROUST) 71
the impression of being constantly sUounded b our consciousness
lam] , it is not as b an unmovable prison; much rather, we feel
carried b it in a perpetual impulse to move beond itself and to
reach outside . . . " (p. 86, II. 39-4) . The epistemologcal sig
nifcance of this impulse is clearly stated when, a few paragraphs
earlier, we heard of a "central belief . . . that made ceaseless mo
tions from inside outward, toward the discover of truth" (p. 8,
II . 36-37) . Like Albertine, consciousness refuses to be captive and has
to take fight and move abroad. This reversal by which the intra
textual complementarit chooses to submit itself to the test of truth
is caused by "the projection of all the forces of life. "
Proust's novel leavs no doubt that this test must fail ; number
less versions of this failure appear throughout the pages of the R
chch. In this section, it is stated wthout ambiguit: "We tr to fnd
again, in things that have thus become dear to us, the refection that
our consciousness [am] has projected upon them; we are disap
pointed in discovering that, in their natural state, the lack the
seduction that, in ou imagination, the owed to the proximit of
certain ideas . . . " (p. 87, II . 2-7) . Banal when taken by itself, the
obseration acquires considerable negative power in context, when
one notices that it occurs at the center of a passage whose thematic
and rhetorical strateg it reduces to naught . For if the "proximit
between the thing and the idea of the thing fails to pass the test of
truth, then it fails to acquire the complementar and totalizing
power of metaphor and remains reduced to "the chance of a mere
association of ideas. " The co-presence of intra- and extra-textual
movements never reaches a sythesis. The relationship between the
literal and the fgal senses of a metaphor is always, in this sense,
metonymic, though motivated b a constitutive tendenc to pretend
the opposite.
The image of the iridescent fountain is a clar case in point .
Everthing orients the trope towards the seduction of metaphor: the
sensor attractiveness, the context, the afective connotations, all
cooperate to this aim. As soon however as one follows Proust's ow
fgr. Grard Gnette quote i t, and it has since been much commented upon,
wthout however exausting the connotations of its context and of its troplogcal
signifcance. Walter Bnjamin well perceived the importance of thi metonyy when
he compared Poust's fgures to a rolled-up sock which is its ow outside and which,
when unrolled, like the Mobius strip, is also its ow inside ("Zum Bilde Proust ,"
Illumintn [Franku am Main, 1955] , p. 30) .
7
RHETORI C
injunction t o submit the reading t o the polarity of truth and error (a
gesture that can be repressed but never prevented) , statements or
strategies that tended to remain unnoticed become apparent and
undo what the fgure seemed to have accomplished. The shimmering
of the fountain then becomes a much more disturbing movement , a
vibration between trth and eror that keeps the two readings from
converging. The disjunction between the aesthetically responsive and
the rhetorically aware reading, both equally compelling, undoes the
pseudo-synthesis of inside and outside, time and space, container
and content, part and whole, motion and stasis, self and understand
ing, writer and reader, metaphor and metonymy, that the text has
constructed. It functions like an oxmoron, but since it signals a
logical rather than a representational incompatibilit, it is in fact an
aporia. It desigates the irrevocable occurrence of at least two mutu
ally exclusive readings and asserts the impossibilit of a true under
standing, on the level of the fguration as well as of the themes.
The question remains whether by thus allowing the text to decon
struct i ts own metaphors one recaptures the actual movement of the
novel and comes closer to the negative epistemolog that would
reveal its hidden meaning. Is this novel the allegorical narrative of its
ow deconstruction? Some of its most perceptive recent interpreters
seem to thin s when they assert , like Gilles Deleuze, the "powerful
unit of the Rchrch despite its inherent fragmentation or, like
Gnette, stress the "solidity of the text" despite the perilous shuttle
between metaphor and metonymy.
lS
Wat is at stake is the possibilit of including the contradictions
of reading in a narrative that would be able to contain them. Such a
narative would have the universal signifcance of an allegor of
reading. As the report of the contrdictor interference of truth and
error in the process of understanding, the allegor would no longer
be subject to the destructive power of this complication. To the
extent that it is not itself demonstrably false, the allegor of the play
of truth and falsehood would ground the stabilit of the text .
One would have to untie the complex interlacing oftruth and lie
in A l rchrh du tmp pedu to decide whether or not the work
corresponds to this model. But the passage on reading gves a frst
18. Gilles Deleuze, "AntiIogos," in Put et l sig, 2d ed. ( Pais, 1970) , and
Genette, "Metonyie chez Poust ," p. 6.
READI NG ( PROUST) 7
indication how such an analysis would have to proceed. It is pre
ceded by an episode (p. 80, 1. 18 to p. 82, 1. 24) which deals, as b
coincidence, with the question of allegor and which can sere as a
warning for the difculties that ay attempt to reach an inclusive
allegorical reading of the novel are bound to encounter. The passge
consists of Marcel's meditaton on the nicknae "Giotto's Chat b
which Swann is accustomed to refer to the kitchen maid persecuted
wth such crelt b Fran<oise, the cook.
Slave of a slave, pathetic emblem of sertude, the kitchen maid
is frst described as what one could call , with Gethe, Daur im
Wechel, the element that remains permanent in the midst of
change. She is characterized as "a permanent institution, whose
unchanging attributes guaanteed an appearance o continuit and
identit, beyond the succession of transitor forms in which she was
incarnated . . . " (p. 80, 11. 25-28) . Swann, the personifcation of
metaphor, is endowed with a particular knack for the discover of
resemblances, and he has obsered the near-emblematic qualit
of this particular ktchen maid. She caries the "humble basket" of her
pregnanc in a manner that, by its resemblance to the surcoat of the
allegorical frescoes painted by Giotto in the Arena of Padua, reveals
her universal essence. All the agonies and all the humiliations of the
successive ktchen maids ae concentrated in this paticular trait of
her physiognomy, thus raised to the level of an emblem. An allegor
thus conceived is in no way distingished from the structure of
metaphor, of which it is in fact the most general version. In the same
manner, metaphor warrants the identit of art as a "permanent
institution" that transcends the singularit of its particular incara
tions. What may appear surprising is that Proust selected sertude
as the essence intended and reached b the fgure. More surprising
still, the allegorical fgure that Swann's sagcit has singled out is
Charit, a virtue whose relationship with sertude is not one of mere
resemblance. By generalizing itself in its own allegor, the metaphor
seems to have displaced its proper meaning.
Marcel , who has a more literar (that is to say, rhetorcally less
naive) mind than Swann, has obsered that the kitchen maid and
Giotto's Charity resemble each other in still another way than physi
cal shape. Their resemblance also has a di mension linked to reading
and understanding, and in this capacit it is a curiously negative one.
The propert shared by the maid and b Charit is that of a nonun
derstanding: both distinguish themselves b features they display
74
RHETORI C
"wthout seeming t o understand their meaning. " Both seem t o be
condemned to the same dyslexa.
The passge desribes with geat precision this shared inability
to read. The allegorical image or icon has, on the one hand, a repre
sentational value and power: Charty represents a shape whose phys
ical attributes connote a certain meaning. Moreover, it maes ges
tures or (in the case of a verbal icon that would no longer be picto
rial) it tells tales that are particularly conspicuous in their intent to
conve meaning. The fgres have to be endowed wth a semantic
intensit that confers upon them a particularly efective repre
sentational function. The allegorical icon must attract attention; its
semantic importance must be dramatized. Marcel insists that the
kitchen maid and the Giotto frescoes resemble each other by their
common claim to fous our attention on an allegorical detail : "Envs
attention-and, b the same token, our ow-[is] entirely concen
trated on the action of her lips . . . " just as "with the poor ktchen
maid, [one's] attention is ceaselessly brought back to her belly b the
load that weighs it down . . . . " In a metaphor, the substitution of a
fgural for a literal designation engenders, by snthesis, a proper
meaning that can remain implicit since it is constituted by the fgure
itself But in allegor, as here described, it seems that the author has
lost confdence in the efectiveness of the substitutive power gener
ated by the resemblances: he states proper meaning, directly or by
way of an intra-textual code or tradition, b using a literal sign which
bears no resemblance to that meaning and which conves, in its
turn, a meaning that is proper to it but does not coincide wth the
proper meaning of the allegor. The facial expression of the "hear
and mannish" matron painted b Giotto connotes nothing charitable
and even when, as in the case of Enr, one could perhaps detect a
resemblance between the idea and the face of Enr, the stress falls
on an iconic detail that sidetracks our attention and hides the poten
tial resemblance from our eyes.
The relationship between the proper and the l iteral meaning of
the allegor, which can be called "allegoreme" and "allegoresis" re
spectively (as one distinguishes between "noeme" and "noesis") , is
not merely a relationship of non-coincidence. The semantic disso
nance goes further. By concentrating the attention of Env's beholder
on the pictuesque details ofthe image, he has , says Marcel, "no time
for envious thoughts. " Hence the didactic efectiveness of allegor
since it makes one forget the vces it sets out to represent-a little as
READI NG ( PROUST) 75
when Rousseau pretends to justif the theater because it distracts, for
a while, vile seducers from their evil pursuits. 1 9 It actually turns out
that , in the case of Env, the mind is distracted towards something
even more threatening than vce, namely death. From the strctual
and rhetorical point of view, however, all that matters is that the
allegorical representation leads towards a meaning that diverges
from the initial meaning to the point offoreclosing its manifestation.
In the case of the allegorical fgation of Charity, thing are
even more specifc, especially if one takes the origins of the passage
into account . Proust does not start out from a direct encounter with
Gibtto's frescoes , but from Ruskn's commentar on Giotto's Vices
and Virtues of Padua. 20 The commentar is of considerable interest
in many respects but it is especially striking in this context because it
deals with an error of reading and interpretation. Rusn describes
Charity brandishing, in her left hand, an object that look like a
heart ; he frst assumes that the scene represents Gd givng his ow
charitable heart to her, but he corrects himselfin a later note: ''There
is no doubt that I misread this action: she gve her heart to Gy
while she makes oferings to mannd. "21 Ruskin also discusses the
painter's ambivalent rhetoric, which is, he says, "quite literal in [its]
meaning as well as fgurative. " Describing the same gestue, Marcel
follows Ruskn's rectifed reading but displaces the meaning by add
ing a comparison which, at frst sight , appears quite inconguous:
"she stretches her incandescent heart towards G or, better, she
hands it over to him,22 as a cook would hand a corkscrew through a
window of her basement to someone who ask for it at street-level"
Cp. 81 , 11 . 22-25) . The comparison seems to be chosen merely to
19. Preface a Nari, inJ. J. Russau, Ouve cmpl, ed. Bernar Gage
bin and Marcel Ryond (Pa: Gllirard [Bibliotheque d la Pleiade], 191) ,
2: 973.
20. On this question, seJ. TheodoreJohnon, Jr. , "Proust and Giotto: Fouda
tions of an Allegorcal Interretation of A l rhh d tmp pru, " in Marel
Put: A Crt Panor (Urbana, Ill . , 1973). That Mr. Johnn's and my concep
tions of alegor have little in common is clea from hi inistence, agint textu
evidence, on "the prfect blending o realit and smbl in the ccle of the Virte
and Vices" Uohnsn, p. 202) .
21 . John Rskn, For Clvgra in Th Work q]ohn Rkin, ed. E. T. Cok
and A. Wedderbur ( Lndon, 1907) , 27: 130.
22. The French text sys "elIe Ie lui pa . " with "pass" italicized, which
sugests varou collouial assoiations. For ou puross, one can conne oneelf to
the connotativ fi eld sugested by the "lowly" implications of the ter.
76
RHETORI C
stress the homely qualit of the gesture, but one ofits other functions
is to bring about the re-entr into the text of "the cook," that is to sy,
Francoise. The kitchen maid resembles Giotto's Charity, but it ap
pears that the latter's gesture also makes her resemble Francoise. The
frst resemblance is not entirel unlikely: the suferings of the hapless
girl are vvidly enough evoked to inspire a feeling of pit that could
easily b confused wth charit. But the further resemblance, wth
Francoise, is harder to understand: if the image, as a representation,
also connote
s
Francoise, it wdely misses its mark, for nothing could
be less charitable than Francoise, especially in her attitude toward
the kitchen maid. The neighboring episode (pp. 120-24) , which nar
rates in great detail the refnements of Francoise's methods of tor
ture, makes ver clear that the literal sense of this allegor treats its
proper sense in a most uncharitable manner. The rhetorcal interest
of the section, which culminates in the tragicomic scene where Fran
coise is seen weeping hot tears upon reading, in a book, a description
of the ver syptoms that prompt her most savage violence when
she literally encounters them in her slave, is that a single icon engen
ders two meanings, the one representational and literal , the other
allegorical and "proper," and that the two meanings fght each other
wth the blind power of stupidit. With the complicit of the witer,
the literal meaning obliterates the allegorical meaning; just as Marcel
is by no means inclined to deprve himself of Francoise's serices, so
the writer has no intention of doing wthout the thematic pwers of
l iteral representation and, moreover, would not be able to do so ifhe
tried.
In the ethical realm of Virtue and Vice, the ambivalences of the
allegorical fgre thus lead to strange confusions of value. And if one
bears in mind that, in Proust's allegor of reading, the couple Fran
coiseltchen maid also enacts the polarit of truth and falsehoo,
then the epistemologcal consequences of the passage are equally
troubling. Since any narrative is prmarily the allegor of its own
reading, it is caught in a difcult double bind. As long as it treats a
theme ( the discouse of a subject, the vocation of a writer, the con
stitution of a consciousness) , it will always lead to the confrontation
of incompatible meanings between which it is necessar but impos
sible to decide in terms of truth and error. If one of the readings is
declared true, it wll always be possible to undo it by means of the
other; if it is decreed false, it will always be possible to demonstrate
that it states the truth of its aberration. A interpretation of A l
READI NG ( PROUST) 77
rehrch du tmp peru which would understand the book as being
the narrative of its ow deconstruction would still operate on this
level . Such an interretation (which is indispensable) accounts for
the textual coherence postulated by Gnette, Deleuze, and by Mar
cel's own critical theories and, at the far end of its successive nega
tions, it will recover the adequation between structure and statement
on which any thematic reading depends. But when it is no longer a
matter of allegorizing the crossing, or chiasmus, of two modes of
reading but Reading itself, the difcult brought to light by the
passage on Giotto's Charit is much greater. A literal reading of
Giotto's fresco would never have discovered what it meant, since all
the represented properties point in a diferent direction. We know
the meaning of the allegor only because Giotto, substituting writing
for representation, spelled it out on the upper frame of his painting:
KA. We accede to the prper meaning b a direct act of read
ing, not by the oblique reading of the allegor. This literal reading is
possible because the notion of charit, on this level of illusion, is
considered to be a referential and empirical experience that is not
confned to an intra-textual system of relationships. The same does
not apply to the allegorical representation of Reading which we now
understand to b the irreducible component of any text. Al that will
be represented in such an allegor wll defect from the act of reading
and blok access to its understanding. The allegor of reading nar
rates the impossibilit of reading. But this impossibilit necessarily
extends to the word "reading" which is thus deprived of any referen
tial meaning whatsoever. Proust may well spell out all the letters of
LCTIO on the frames of his stories (and the novel abounds in
gestures aimed in that direction), but the word itself wll never
become clear, for according to the laws of Proust's own statement it
is forever impossible to read Reading. Everthing in this novel sig
nifes something other than what it represents, be it love, conscious
ness, politics, art, sodomy, or gastronomy: it is always something else
that is intended. It can b show that the most adequate term to
designate this "something else" is Rading. But one must at the same
time "understand" that this word bars access, once and forever, to a
meaning that yet can never cease to call out for its understanding.
The young Marcel is at frst displeased by the discordance be
tween the literal and the proper meaning of the allegor, but the
maturity of his literar vocation is dated by his abilit to come to
admire it: "Later on, I understood that the uncanny attraction, the
78
RHETORI C
s
p
cifc bauty of these frescoes was due t o the prominent place
taen up by the sybol, and that the fact that it was not represented
smbolicaly (si

ce the sbolized idea was not expressed) but as


something real , actually exprienced or materially handled, gave to
the meaning of the work something more literal and more precise
. . . " (p. 82, 11. 7-14) . This formulation, "plus tard, j'ai compris," is
ver familiar to readers of the Rhrch, for it punctuates the entire
novel lie an incantation. Literar criticism has traditionally inter
preted this "later on" as the moment of flflment of the literar and
aesthetic vocation, the passage from experience to wting in the
convergence of the narrator Marcel wth the author Poust . In fact,
the unbridgeable distance between the narator, allegorical and
therefore obliterating fgre for the author, and Poust , is that the
former can believe that this "later on" could ever be located in his
own past . Marcel is never as far away from Proust as when the latter
has him say: "Happy are those who have encountered trth before
death and for whom, however close it may be, the hour of truth has
rung before the hour of death. "23 As a witer, Poust is the one who
kows that the hour of trth, lie the hou of death, never arrives on
time, since what we call time is precisely truth's inabilit to coincide
with itself. A l rechrch du tmps perdu narrates the flight of
meaning, but this does not prevent its own meaning from being,
incessantly, in fight.
23. 3:910.
4 Genesis and Genealogy
(Nitzh)
I N LI TERARY STUDI ES , STRUCTURES OF MEANI NG ARE FRE
quently described i n historical rather than i n semiological or rhetor
cal terms. This is, in itself, a somewhat surprising occurrence, since
the historcal nature of litera discourse is by no means an a prr
established fact , whereas all literature necessarily consists oflinguis
tic and semantic elements. Yet students of literature seem to shy
away from the analysis of semantic structues and feel more at home
wth problems of psycholog or of historiography. The reasons for
this detour or flight from language are complex and go far in reveal
ing the ver semiological properties that are being circumvented.
They explain the methodological necessity of approaching questions
of literar meaning by ways of the nonlinguistic referential models
used in literar histor. This is one of the means, among others, to
gain access to the enigmas that lie hidden behind the more tra
ditional problems of literar classifcation and periodization.
One recurring such problem is that of the genetic pattern of
literar histor and of literar texts that are assumed to reflect, by
analog or by imitation, this pattern-as in the narrative shape of
stories that also purport to be histories (Th Histor ojTom]on, La
Vie d Maranne) , or of histories so neatly framed that the seem to
consist of a single narrative unit (From Baulaire to Surralism,
From Clsi to Rmnti) . The question is prevalent in attempts at
self-defnition with relation to the past , as when the contemporar
mind is said to be, for instance, "post-Rmantic" or "anti-idealist . "
Rmanticism itself is generally understood as the passage from a
mimetic to a genetic concept of art and literature, from a Platonic to
a Hegelian model of the universe. Instead of being mere copies of a
transcendental order, Nature or G, "all things below" are said to
be part of a chain of being underay to its teleological end. The
hierarchical world of Ideas and Images of Ideas becomes a world of
means movng towards an end and ordered in the prospective tem-
80
RHETORI C
poralit ofa genetic movement . The exstence of this en justifes the
claim of the Romantics not to be a mere repetition of former perfec
tion but a true birth, a beginning. "Das Rsultat ," sys Hegel , "ist nur
darum dasselbe, was der Anfang, weil der Anang Zweck ist" (Ph
nomlog o th Mind, Introduction) . The English translation of
the words italicized by Hegel illustrates the interdependence and
potential identit of end and beginning that characterizes a genetic
concept of time: "The outcome is the same as the beginning ony
because the beginning is an end. "
It would be tempting to document the emergence of the genetic
pattern within the Romantic imagination and Romantic rhetoric.
The prevalence of this pattern is not yet understod in all its implica
tions and many studies of Rmanticism are still in a pre-Hegelian
stage. The tradition is caught in a non-dialectical notion of a
subject-object dichotomy, revealing a more or less deliberate
avoidance of the moment of negtion that coincides, for Hegel , with
the emergence of a true Subject . Such a study could lead us far in
undoing a system that puts a natural , organic principle at the center
of things and constructs a series of analogical emanations around
this center, ending up wth an altogether un-Hegelian concept of the
subject as an irational, unmediated experience of prticular self
hood ( or loss of selfood) . It would show that a dialectical concep
tion of time and histor can ver well b genetic and that the
abandonment of an organic analogism by no means implies the aban
donment of a genetic patter. Wen a contemporar philosopher like
Michel Foucault characterizes nineteenth-centur late-Rmantic his
toricism as "lodged within the distance between particular histories
and universal Histor, between singular events and the Orign of all
things, between evolution and the frst division within the source,
between forgetting and return,'" then the vocabular of source, ori
gin, distance, memor, indicates that we are more than ever dealing
wth a genetic model defned in terms of an intent oriented towards
an "end. " The allegorization and ironization of the orgnic model
leaves the genetic patter unafected.
It also leaves unafected the genetic structure of the historiog
raphy that deals wth Rmanticism i tself, as it developed during the
nineteenth and the twentieth centur. Within an organically deter
mined vew of literar histor, Rmanticism can appear as a high
1 . L Mot et l ch ( Paris, 19) , p. 231.
GENESI S AND GENEALOGY ( NI ETZSCHE)
81
point, a period of splendor, and the subsequent centur as a slow
recedig of the tide, a decay that can take on apocalyptic pro
portions. A reversed image of the same moel sees Romanticism as a
moment of extreme delusion from which the nineteenth centu
slowly recovers unt it can free itselfin the assertion of a new moer
nity; Nietzsche himself, violently anti-Rmantic in his cultual
ideolog, invariably adopts that perspective when he writes organic
or, in his terminolog, monumntal histor. The critical "deconstruc
tion" of the organic moel changes this image: it creates radical
discontinuities and disrupts the lineait of the temporal process to
such extent that no sequence of actual events or no paticular subject
could ever acquire, by itself, full historical meaning. They all become
part of a process that they neither contain nor reflect , but of which
the are a moment . The can never b the source or the end of the
movement , but since the movement consists of their totali zaton,
they can still be said to share in the experience of this movement. No
father, no son can be G, but the histor of the strgle between
fathers and sons remains in essence divine. As a diachrony animated
by a teleological intent, such a movement remains genetic. The in
tentional principle is no longer some ideal model or hypostasis but
the law of the ultimate conformit of the end to' the origin. Any
particular subject or event , including texts, can b ordered as a mo
ment within this conformit; this interpretative act of orderig and
of classifcation both understands the event and locates it within the
diachrony of the movement . In such a system, histor and interpreta
tion coincide, the common principle that mediates between them
being the genetic concept of totalization.
To wite a histor of Romanticism that would no longer b
organic but still genetic would be ver useful , all the more since no
trly dialectical histor of Romanticism has as yet been witten.
Hegel's outlines ofliterar or at histor bypass, as is well kown, the
contemporar moment entirely and this prdictable blindness is re
peated in later work that a the products of genuinely dialectical
minds, such as Auerbach's Mimi or Walter Bnjamin's Ursprng
d dutchn Traurpil. But the question remains whether such a
dialectical histor of Rmanticism could do justice to its object. Can
the genetic pattern be said to be "trly" chaacteristic of Romanti
cism? Does this system, wth all the conceptual categories that it
implies ( suect, intent , negation, totalization, supported by the un
derlyng metaphysical categories of identit and presence) remain as
8
RHETORI C
uncha
llenged i n writers of the late eighteenth centur as i t remains
unchallenged in most of their later interpreters? It could b that the
so-called Romantics came closer than we do to undermining the
absolute authorit of this system. If this were the case, one may well
wonder what kind of historiogaphy could do justice to the phenom
enon of Rmanticism, since Rmanticism (itself a period concept)
would then b the movement that challenges the genetic principle
which necessarily underlies all historical narrative. The ultimate test
or "proof of the fact that Rmanticism puts the genetic patter of
histor in question would then be the impossibili
t of writing a
histor of Romanticism. The abundant bibliogaphy that exists on
the subject tends to confrm this, for a cuious blindness seems to
compel historians and interpreters of Romanticism to circumvent
the central insights that put their ow practice, as historans, into
question.
One way of progressing in this difcult question involves the
examination of texts which, by their ow structure and their own
statement , lay the foundation for the genetic conception of histor.
From the eighteenth to the ver recent twentieth centur, one could
select from a wide varie
t
of such texts, from Montesquieu's Eprit
ds loi to Russeau's Diours sur l
'
origne d l'inalit, from Hegel's
Phomnolog ofth Min to Heideger's "The Origin of the Work
of Art" ("Der Ursprung des Kunstwerkes," in Holzwee) . Even such
recent exmples as Michel Foucault' s or Jacques Derrida's attempts
to see the conceptual crisis oflangage that fgures so prominently in
contemporar philosophy, as closing of a historical priod, some
times specifcall designated as the "epoque de Russeau," fall within
this pattern.
The choice of Nietzsche' s Th Birth ofTraged as a text particu
larly well suited for this purpose needs little justifcation. Within the
system of historical periodization implied by an "epque de Rus
seau," Nietzsche represents an important articulation. Together with
Marx and Freud, in a triumvrate that has become a cliche ofintellec
tual histor, his work participates in the radical rejection of the
genetic teleolog associated with Rmantic idealism. Within the cor
pus of his own work, the pattern is repeated in the development that
is said to lead from the early Birth ofTragedy ( 1871) to the entirely
diferent tone and manner that prevails, in the published work, from
Human all too Human ( 1876-78) , on. In its ow structure as well as
in its historical function, his work would be a critique ofthe Rman-
GENESI S AND GENEALOGY ( NI ETZSCHE) 8
tic ideolog, concluding the perod that can b said to start wth
Russeau. And it would indeed be difcult to fnd a text in which the
genetic pattern is more clearly in evdence than in Th Birth of
Tragedy: it operates on various levels that all spring from a common
source and converge toward a common end. We can take time to
examine only some of these levels, but this should sufce to give this
exercise in genetic "deconstrction" a more than heuristic sig
nifcance.
TIl Birh ofTagedy is rgtly considered to be one of Nietzche's
most unifed texts. "It would be, in the fnal analysis, Nietzsche's only
genuine 'Book,' " says Philippe Lacoue-Labarthe,
2
in a comment that
fails perhaps to do justice to the coherence of A Genalog ofMoral,
but that still faithfully refects the frst impression of any Nietzsche
reader. Compared to the near-contemprar Philosophnbuh3 or to
the subsequent Human all too Human, Th Birth of Tragedy seems to
defend a well-rounded thesis, supported by relevant argument and
illustration. As the title sugests, the principle of this coherence is
unquestionably genetic, a classical example of the mode: the histor
of a birth and a rebirth, lie Dante's Vita nuova but also like
Nietzsche's favorite novel, Ttrm Shandy. The text is held together
by the psuedo-polarit of the Apollo/Dionysos dialectic that allows
for a well-ordered teleolog, bcause the ontological cards have been
stacked from the beginning. With unquestionable fdelit to the
dyamics of the text, Gilles Deleuze can say: "In a tragedy, Dionysos
is the tragic essence [I fond du tragu] . He is the only tragic
character, the 'sufering and glorifed deit; his sufering are the
only tragic subject , the suferng of indivduation reabsorbed in the
joy of original oneness."
4
Wat is being said here about tragedy in
general would seem to apply to Nietzsche's text as well. Truth, Pres
ence, Being are all on Dionysos's side, and histor can only occur as
2. "L Dtou," Pou 5 ( 1971) : 52.
3. A sres of fragments and aphorisms that tiled to coalese into a com
plete b k and now appar in Volume 6 of the s-called Musron edition of
Nietzshe's complete work (Friedrch Nietzshe, Gmmlt We [Munich:
Musrion Verlag, 1920]). Page numbr of quotations from T Birh oTgy are
from the Musron edition, Gmmlt we, Volume 3. The numbrs that follow
the volume and page numbrs refer to the numbred sctions in which the text i
divded.
4. Niet:h e 1 philphie (Pari, 192) , p. 13.
8 RHETORI C
the birth and rebirth of a father i n whose absence no son could ever
exist. The starting point, Dionysos, contains wthin itself the end
point, the Apollonian work of art, and governs the dialectical path
way that leads from the one to the other. Any cross-section made in
the diachrony of the histor can be valorized in terms of the greater
or lesser manifestation or presence of Dionysos, the original
"ground" by means of which distance and proximit can be mea
sured: Sophocles is glorifed, Plato and Euripedes cast as near-villains
because of their greater or lesser proxmit to Dionysos. The same
criteria apply in the modern period, in the criticism of Florentine
opera, of imitative music, and of the modern drama, or reversely, in
the extravagant claims made for Wagnerian opera.
The imager of depth and foundation used by Deleuze to convey
the priorit of Dionysos receives support from many statements
throughout the text. It is less dependent, however, on the diachronic
narrative than is the case in other genetic work such as Rousseau's
Dicourse on Inequlit or his Esay ofth Orign ofLangge. In these
texts, the narrative articulations become themselves important
thematic categories and the genetic moments are presented as cos
mic catastrophes or divine interventions. By contrast, the outward
narative transitions in Th Birth ofTragedy ofen consist of mere
formal symmetries devoid of thematic weight. Thus the rebirth of
Dionysos in the person of Wager, crcial as the event may be, is
described as a mere reversal of the regressive movement that de
stroyed the Hellenic world into a symmetrical movement of regener
ation by which the xodern, Grmanic world is to be reborn.5 Pas
sages of this kind are valueless as argments, since they assume that
the actual events of histor are founded in foral symetries easy
enough to achieve in pictorial, musical, or poetic fctions, but that
can never predict the occurence of a historical event. The narrative
links are s weak that one may feel tempted to put the unit of the
text in question for purely philological reasons, on the grounds that
5. "If we have rigtly . . . linked the disppearance of the Donysian spirit to
an obvious but still unexplained transformation and degeneration of the Greek,
what hopes should not be kndled in us when we obsere unmistakble auspices that
th re pres, th gaul rwakning Of th Dinyin spirt, is tang place
in ocr contemporar world! " (Birh Of Tae, 3: 133; 19) "we are reexperiencing
analogically the geat Hellenic priods in red ordr . . and for example now
seem to b moving backwards from the Aexandrian age into the pero of traged"
(ibid, 3: 135; 19).
GENESI S AND GENEALOGY ( NI ETZSCHE) as
the complex inception of Th Birh afTagedy maes the fnal prod
uct into a patchwork of disconnected fragents or, as Nietzsche
himself put it , into a "centaur. "6
The relative weakness ofthe narative coherence becomes much
less important when one realizes that the diachronic, successive
structe of Th Birh ofTagey is in fact an illusion. We normally
thin of genetic patters as successive in time, and Th Birh of
Taedy indeed ofen describes the Apollonian and Dionysian phases
as "always new successive births" (3:39; 4) . It can be shown, how
ever, that whenever an art form is being discussed, the three modes
represented by Dionysos, Apollo, and Socrates are always simulta
neously present and that it is impossible to mention one of them
without at least implying the others . The Dionysian moments always
occur in revolt against the tanny or as a result of the failure of the
Socratic claim to knowledge; the Dionysian insight must always be
doubled at once b the Apollonian shelter of appearances; and the
Apollonian vision is always the vision of "the eternal contradiction, of
the father of all things" (3:37; 4) . Yet this simultaneit does not
disprove the persistence of a genetic model , since parental relation
ship can be described as synchronic structures without in the least
denying their genetic nature. As long as the Dionysos/Apollo rela
tionship is referred to, as in the previous quotation, in an imager of
parenthood, successiveness and simultaneit are in fact mirror-like
versions of the same ontological hierarchy. And although the struc
ture of parental imager in Th Birh afTagedy is inconsistent, the
metaphors nevertheless remain familial throughout .
The genetic structure ofthe text i s confrmed by layers of mean
ing that are sturdier than the formal symetres of the narrative
plot . Other genetic linages are at play, ofen based on genuine
philosophical insight rather than on the manipulation of geometrical
metaphors. Thus the transition to Wagerian modernit fnds its
thematic equivalence in the movement from science to art , from the
most extreme forms of epistemologcal constraint to the liberating
infuence of Grman music. The myth of Socrates mas this move
ment into a historical development since Socrates, undoer of Greek
tragedy and founder of modern epistemolog, rpresents the deca
dence that a new modernity has to overcome. The transformation of
6. Ltter to Rohde, Febrar 1870, Br 2; 3: 183, also quoted in the editor's
pstface [Nachbrcht] to T Birh ofTagey, Musron, 3:401 .
8 RHETORI C
the epistemologcal into an aesthetic model i s not to be thought of as
a mere value assertion, an uncritical preference for the irational
rather than the rational facuties of man. The relationship beteen
science and art is a great deal more complex from the start and,
already in Th Birh of Tagey, Nietzsche advocates the u of epis
temologcaly rgorous methods as the only possible means to refect
on the limitations of these methods. One cannot hold against him
the apparent contra
d
iction of using a rational mode of discourse
which he, in fact, ner abndoned-in order to prove the inadequa
c of this discou. At the time of Th Bir qTey, Nietzshe is
entirely in control of this problem and can state it with full thematic
clarit, precisely in describing the transformation of the Socratic into
the Wagnerian man. He uses and remains faithful to the Kntian
element in Shopenhauer's terminolog and this allegance is itself
epistemologcally founded: "geat men, capable of truly general in
sight, were able to use the devces of science itself in order to reveal
the limits and relativt of all knowledge, thus decisively putting into
question the scientifc claim to universal validit and purpose. Their
demonstration undid for the frst time the illusion that the essence of
things can be reached by means of causalit (3: 123; 18) . Thes
"gat men" are identifed as Knt and Shopenhauer, and the refer
ence to the laws of causalit orients the remark towards the most
rigorous epistemological sections of the Crtu of Pure Ran. The
sme strateg and progammatic outline for the later work, in which
the word "science" will reappear with a positive valorization (as in
the title Th Gay Sine) , is contained in a fragent that dates from
the same perod as Th Birh qTaedy: "Control over the world by
means of positive action: frst through science, as the destroyer of the
illusion, then throug art, as the only remaining mode of existence,
because it cannot be dissolved by logc" ( Musarion, 3:212, Frag
ment 4) .
In Th Birh of Taey, the problem ofthe relationship beteen
art and epistemolog functions within a genetic as well as a
philosophical perspective: the ambivalence of the epistemological
moment, as the critical undoer ofits own claim at universl veracit,
is represented as a genetic development from the Aexandrian to the
truly moern man, and undoubtedl owes some ofits persuasiveness
to the narrative, sequential mode of presentation. It does not how
ever, in this text, put this mode explicitly or implicitly into
question. The genetic structure of Th Birh of Tagedy is not af-
GENESI S AND GENEALOGY ( NI ETZSCHE) 87
fected, literally or rhetorically, by the epistemological paradox that it
contains. Instead, the logical complication can rightly b said to b
one of the principal articulations of the narrative. We must look
further in order to discover whether the genetic pattern of Th Birh
of Tragedy is substantial ( Le. , motivated by thematic statement con
sidered as maning) or rhetorical (Le. , motivated by thematic state
ment considered as struture) .
There may well be an underlyng, deeper pattern of valoriztion
that confers genetic coherence and continuit upon the text and that
transcends the thematics of the Apollo/Dionysos or the Scrates/
Dionysos dialectic. A great deal of evdence points to the likelihoo
that Nietzsche might be in the grip of a powerful assumption about
the nature oflanguage, bound to control his conceptual and rhetori
cal discourse regardless of whether the author is aware of it or not .
By the ver choice of its literar theme, The Birh of Tragedy seems
concerned, ofits ow volition, with what a text is or ougt to be. Yet ,
in the fnal version, little i s explicitly being said about the nature of
literar language. We hear a lot about various subjects, mostly histor
ical and cultural : a not entirely orgnal or respectable theor about
the importance of the chorus in Greek tragedy; considerations on the
parallel rise and fall of the Greek theater and the Greek state and on
the use and abuse of the sciences and of philolog in contempora
education; agresively polemical attack on certain art forms wrtten
with a curious rancor; a highly personal and exalted plea for Wagner
wth strongly nationalistic overtones-all enclosed wthin the general
framework of a mythological narrative involvng two entities,
Dionysos and Apollo, that are explicitly said to exist on a purely
physiological as well as on a linguistic level. Throughout the main
text and the preparator fragents, the importance of language is
consistently undercut: we are told that we can have no idea what
Greek tragedy was like in the absence of the nonverbal components
ofthe performance; in listening to vocal music, the text is an obstacle
to the pure sound of the voice and it should never be understoo in
the frst place; the outline for an elaborate theor of the work of art ,
reminiscent of the speculations of another post-Wagerian, Mal
larme, is entirely directed towards the suppression of text in favor of
mime and symphonic music. When the topic itself requires that
literar langage be considered, the chosen art form, Greek tragedy,
caries with it such a weight of ideological , cultural , and theologcal
experience, that it is nearly impossible to work one's way back to
8
RHETORI C
linguistic elements. The main theoretical speculations on language
and art that originated at the time of Th Birh of Tagedy have not
been included in the fnal version. Nietzsche alludes to this in a letter
to Rohde that accompanies the manuscrpt of the lecture "Scrates
and Tragedy" : "a curious metaphysics of art , which seres as back
gound [to the main text of Th Birh of Traey] , is more or less my
property, that is to say real estate [Grnbeitz] , thoug not yt
circulating, monetar, and consumed propert. Hence the 'puple
darkness,' an expression that pleased me more than I can s (post
face, 3:401) .
What is this "real estate" that supports the phantasm, the
rhetorical currenc that the fnal version of Th Birth of Traedy puts
in circulation? It must be something more fundamental than the
dialectical interplay of Dionyos and Apollo, since the polarities
would themselves be rooted in it. Everthing seems to sugest that
this "property" stems indeed from the dispossesion of the word in
favor of music. The property rights over truth that belong, by
philosophical authority, to the power of language as statement, are
transferred to the power of language as voice and melody. Jacques
Derrda has identifed this gesture, which he calls "logocentrc," as
the perennial movement of all metaphysical speculation and has
traced some of its versions in Plato, Hegel , Rousseau, Heideger,
Saussure, and others. Far from weakening the grounding of
philosophy in ontolog and in a metaphysics of presence, the transfer
that favors voice over writing, art over science, potr over prose,
music over literature, nature over culture, symblical over concep
tual language (the chain of polarities could be extended at length and
could also be put in a less naive terminolog) , seres in fact to
strengthen the ontological center (theocentric, melocentric, logocen
tric) and to refne the claim that truth can be made present to man.
It also recovers the possibility oflanguage to reach full and substan
tial meaning. A great number of passages from Th Birth of Trgey
seem to place the text forcefully wthin the logocentric tradition. The
later evolution of Nietzsche's work could then b understod as the
gadual "deconstruction" of a logocentrism that receives its fullest
expression in Th Birh of Taedy.
The logocentric valorization necessarily implies the persistence
of the genetic model as its only possible representation in temporal
or hierarchical terms. The propsitions by means of which Nietzsche
seems to identit himself most forcefully with music are always
GENESI S AND GENEALOGY ( NI ETZS CHE) 89
stated in terms of genetic fliation, as when he compares the absur
dity of having music originate out of potr to the absudity of having
the father originate out of the son ("Tring to illustrate a poem by
means of music . . . what an upside-dow world! A procedue that
strikes me as if a son wanted to sire his father! " [3: 343]) . Within the
reading of Th Birh of Tagedy as a logocentric (or melocentric) text ,
the relative weaess of the main thematic articulations is of little
importance, since the pattern is rooted in a deep-seated generative
conception oflanguage that is boud to control all the movements of
the work, regardless of whether the belong to the highly self
reflective or to the loosely rhetorical levels of the discourse.
The most recent readings of Th Birth of Tragey are still
oriented in this direction and do not question its logocentric ontol
og. Thus for Philippe Lcoue-Labarthe, who has admirably
doumented the importance of rhetoric for Nietzsche: "one would
fnally have to admit that , at the point which Nietzsche's 'decon
struction' [ i . e. , of the logocentric discourse] is able to reach, we a
still standing under the aegis of trth; the labor of trth goes on,
since we are trng to recall something that has been forgotten, to
reveal something unconscious, to fnd the path of a 'reminiscence. ' "7
And Sarah Kofman states the same position in a more apoictic tone
in her discussion of Nietzsche's concept of metaphor at the time of
Th Birth of Tragedy, as compared to his later concept of 'interpreta
tion." In the early text , the stress on the sybolical natu of the
language is still part of a binar system that opposes metaphorical to
literal meaning and that reasserts, wlly nilly, the authority of mean
ing. ''To conceive of the essence oflanguage as rhetorical implies the
reference to a 'truthful' form oflanguage, and a devaloriztion of the
rhetorical in favor ofliteral language. "
s
Later on, well afer Th Birth
of Traey, this assumption presumably disappears: "the concept of
metaphor becomes entirely un-prper [imprpre] , for i t no longer
refers to an absolute proper meaning but always already to an in
terpretation. " It would seem that, in Th Birth of Tagey, this "abso
lute" or "truthful" meaning is the melocentric Gd Dionysos. The
relationship btween Dionysos and Apollo is again stated, by Sarah
Kofman, in the genetic language ofa father/son relationship and the
7. Lacoue-Labarthe, "L Detour," p. 73.
8. "Nietzsche et la metaphore," Poeu 5 ( 1971 ) : 78.
9
RHETORI C
historical version ofthis relationship reappears in the development of
Nietzsche's later works. The same pattern is always repeated: wthin
Nietzsche' s complete work, in the histor of Romanticism, in the
relationship btween Russeau and Nietzsche, in the relationship
between Romanticism and modernit, etc. We now begin to see what
is at stake in the reading of Th Birth of Taedy, what problem
stands behind the a prr assertion of the genetic strcture: the
relationship between langage and music, beteen literal and
metaphorical diction, between narrative ( diegesis) as representation
and narrative as temporality. Nietzsche was certainly right when he
referred to the nature of the Dionysos/Apollo relationship as "th
capital question [die Hauptag]" (Musarion, 3: 357) .
In movng from a thematic to a more rhetorical reading of Th Birh
ofTaedy, we can take our clue from an explicit statement in which
Nietzsche asserts that the rlationship between music and images
( Dionysos and Apollo) is not comparable to the relationship beteen
body and soul , but that it must be understood as "the opposi
tion between appearance and the thing itself ("Ggensatz der
Erscheinung und des Dinges" [3: 146; 21]) . "Thing" is not just what
we usually call reality but Ding an sih, the entity as substance in its
identity wth itself This terminolog faithfully reproduces the prn
ciple of articulation that functions throughout the text . The succes
sive incarnations ofthe Apollonian and the Dionysian spirit, from the
physiological description of the Apollonian as dream to the highly
evolved form of the Wagerian drama, are always structured in
terms of these categories.
The genetic version of the polarity Appearance/Thing is that of
an entity that can be said to be identical wth itself and that would
engender, through a process of mediation , an appearance of which it
is the origin and the foundation. Such a moel can be understood in
linguistic terms as the relationship between fgral and proper
meaning in a metaphor. The metaphor i not "really" the entit it
literally means, but it can b understood to refer to something in
which meaning and being coincide. The meaning engenders and
determines the metaphor as the appearance or sign of this meaning.
This also seems to be how Nietzsche conceives of metaphor when he
wites: ''The metaphor, for the true pot , is not a rhetorical fgure,
but a substitute image that actually exists for him [d ihm wirklich
vorshJbt] , instead of a concept" (3:60; 8) . The metaphor does not
GENESI S AND GENEALOGY ( NI ETZSCHE) 91
mean what it says but , in the last analysis, it says what it means to
say, since it rmains controlled by and oriented towards a specifc
meaning or set of meanings. Such a concept of metaphor coincides
with the ver notion of langage conceived as a system of symbolic
maning, and Th Birh of Tragedy ofers many convincing reasons
why the detour through the metaphorical realm of appearances is
necessar.
There is little difcult in matching the two mythological ples,
Dionysos and Apollo, with the categories of apparance and its an
ti thesis, or wth the relationship between metaphorical and proper
language. From its frst characterization as dream, Apollo exists
entirely within the world of appearances. The dream is not , in Th
Birh of Tragedy, the emergence of a "deeper" trth hidden by the
distraction of the wakeful mind; it is a mere suface, a mere play of
forms and associations, an imager of ligt and color rather than the
darkess of the "nether Sphere. " Far from being a loss of conscious
ness, it remains persistently aware, even in its "sleep," of its illusor,
fctional character, and it delights in this illusion. It is not a revealed
consciousness, since what it shows was never hidden. And it is not a
false consciousness, since it does not for a moment have the illusion
that its illusions are realit.
It complicates but does not , at least from our point of view , alter
the situation that this state of illusion happns to coincide with what
is usually called "realit in everday speech, the empirical reality in
which we live. In this reality, we must vew ourselves "as the trly
nonexistent , i . e. , as a constant becoming in time, space, and causal
ity, or, in other words, as empirical realit (3:36; 4) . The quotation
comes from the section on the epic and seres to stress the doubly
fantastic quality of all narative realism: not only is it the representa
tion of an event and not the event itself, but the event itself is already
a representation, because all empirical experence is in essence fan
tastic. As mere appearance of appearance [Shin ds Shin] ,
Apollo dwells unquestionably in the realm of appearance.
All appearance, as the concept implies, is appearance of some
thing that , in the last analysis, no longer seems to be but actually is.
This "something" can only be Dionysos. Contrar to the dream, de
void of actuality, the intoxication which is said to be the physiologcal
equivalence of Dionysos takes us back to the origin of things, pre
cisely to the extent that it awakens us from the sleep of empirical
reality. As such, the Dionysian condition is an insight into things as
92 RHETORI C
they are and it reaches truth by a negative road, by revealing the
illusor nature of all "realit. " The Apollonian appearance is the
metaphorical statement of this trth; the actual meaning of the
Apollonian appearance is not the empirical realit it represents but
the Dionysian insight into the illusor qualit of this realit.
If this is the case, then the priorit of the musical, nonrepresen
tational language of Dionysos over the representational, graphic lan
gage of Apollo is beyond dispute. Dionysos bcomes indeed the
father of all art , including the plastic arts. Painting becomes a pre
liminar art form that prefgres truth and only waits for Dionysos to
give it vice. A fragmentar passage not included in the fnal version
states the inevtable triumph of music over painting:
Let us people the air with the fantasies of a Rphael and watch,
as he did, Saint Cecilia listening ecstatically to the choirs of
angels. Not a sound emanates from this world apparently lost in
music. Indeed, if we imagine that , by some miracle, these har
monies suddenly became audible, Cecilia, Paulus, and Mag
delena, even the Heavenly choir, would suddenly disappear into
nothing. We would at once stop being Rphael and, as on this
painting the worldly instruments lie shattered on the foor, our
painter's vision, defeated by a higher power, would pale away
and vanish like a shadow. [Musarion, 3: 343]
A diferent version of the same statement occurs in the description of
Wagnerian opera: although the relationship between Dionysos and
Apollo is then said to be a fraternal equalit. Nietzsche insists, on the
other hand, that Dionysos "is powerful enough ultimately to push the
Apollonian drama into a sphere where it begins to speak wth Diony
sian insight , denyng itself and its Apollonian phenomenalit
(3: 147; 21) . The tlos of this ultimate denial is what matters: seen
from that perspective, all previous alliances between Dionysos and
Apollo, marital , fraternal, paternal , or merely structural, are super
seded by the genetic power vsted in Dionysos as the father of all art,
of all appearances.
Why then, if all truth is on Dionyos's side, is Apllonian art not
only possible but even necessar? Why the need for metaphorical
appearance since the proper meaning is all that counts? The question
can still be answered from wthin the genetic logic of Th Birh of
Tagedy and without havng to undo the pattern of this logic. It
GENESI S AND GENEALOGY ( NI ETZSCHE) 9
follows directly from the characterization of Dionysian insigt as
trag insight. The discover that al empirical reality is illusor is
called a tragc discover; no man, it seems, would be able to with
stand its destrctive power. Nietzsch
e
can only bring us closer to it
by a series of myhologcal approxmations. The one who has
reached it is, like Hamlet , frozen forever in the madness of inaction.
Like Silenus, the best companon of Dionysos, he knows that "what is
best of al lies forever beond you reach: not to have been born, not
to be, to be nnthing. The second best however is for you-to die soon"
(3:32; 3). He realizes, lie Oedipus, that trth can only be reached at
the cost of ultimate moral transgression and, like Pometheus, he
experiences the essentially contradictor nature of the world in a
state of endless rebellion devoid of hope. The entire semi-popular
"existential" reading of Nietzsche takes offrom this particular tonal
ity. All readers of Th Birth ofTagedy know by means of what ruse
the destrctiveness of un mediated trth is avoided: instead of being
directly exprienced it is represented. We are rescued by the essential
theatricality of art. "Only as an athti phnnmn is existence
and the world foreverjutf" (3:46; 5, and 3:261 ; 2) : the famous
quotation, twce repeated in Th Birth of Tagedy, should not be
taken too serenely, for it is an indictment of exstence rather than a
panegric of art . It accounts however for the protective nature of the
Apollonian moment . The Apollonian ligt , says Nietzsche in one of
his most string metaphors, is the mirror image of a well-known
optical phenomenon: "Wen, in a determined attempt to look di
rectly at the sun, we have to turn away blinded, we'll have dark spots
before our ees to shelter them from the sunrays. Convrsely, the
brightly projected images of the Sophoclean hero, the Apollonian
mask, are the necessar consequences of a glance into the inside and
terrors of nature, bright spots to heal eyes wounded by the fearful
nigt" (3:65; 9) .
It is time to start questioning the explicit , declarative statement
of the text in terms of its own theatricality. The system of valoriza
tion that privleges Dionysos as the truth of the Apollonian appar
ance, music as the trth of painting, as the actual meaning of the
metaphorical apparance, reaches us through the medium of a
strongly dramatized and individualized voice. Still more than Rus
seau's Dioure on th Orgn ofInqulit, Th Birth ofTagedy is
indeed a discourse, a harange that combines the seductive power of
a genetic narrative with the rhetorical complicity of a sermon. A
1
RHETORI C
reealing self-crtical statement about his ow literar manner, in
the Philphbuh, speak of Nietzsche's resolution "to wite, in
general , in an impersonal and cold manner. Avoid all mention of , us'
ad 'we' ad ' I. ' Also limit the number of sentences wth relative
clauses" (Musion, 6: 62) . The opposite happens, of course, in Th
Birh o Tae. The complicit between the "I" of the narator and
the collective ''we'' of his acquiescing audience functions relentlessly,
uderscored by the repeated addess of the audience as "my friends. "
The orator has our best interests at heart and we are garanteed
intelectal sfet as long as we remain wthin the shelterng reach of
hi voice. The same seductive tone safegards the genetic continuit
througout the text , easing the listener oer difcult transitions by
means of helpful summaries, marking out the truly important points
by attention-catching signals. The more delirious passages are clealy
mrked of as digressions, afer which "we gide back into the mod
beftting contemplation" (3: 139; 21) , thus gining at once our con
fdence with the reassuring thougt that someone who allowed him
self such verbal excesses in his digessions must be ver cool and
contemplative indeed in his argmentations. The voice is not beyond
crediting us, the raders, with praise that it lavishes upon itself, as
when we are endowed with Dionyian insight in being told that "we
pnetrate, with piercing clarit, into an inner world of motives"
(3: 159; 2) . A longer enumeration of examples is superuous, since
the are as numerous as the are obvious in their strateg: this orator
is in need of a ver benevolent audience if it is to accept a shaky
sstem of valorization.
The need to dramatize emphatically the stance ofthe convinced
man is indeed imperative, for the genetic valoriztion of Dionysos
reintroduces wthin the text all the categories that had orignally
been put into question, including the notion of an ontologically
rooted stem of values. For all its genetic continuit, the movement
of Th Birh of Taey, as a whole as well as in its component parts,
is curously ambivalent wth regard to the main fgs of its ow
discourse: the categor of reprsentation that underlies the narrative
moe and the categor of the subject that supports the all-perading
hortator voice.
Rpresentation (mostly referred to, in this text , as Vorstllung or
Abil) functions throughout wth a negative value-emphasis. From
a purely historical point of view, Th Birh of Taed could be
ordered among the pre-expressionist critical documents in which a
GENESI S AND GENEALOGY ( NI ETZSC HE)
95
nonrepresentational art is being prepared; this may well be the text's
main function in the histor of crticism. Te principal targets of this
critique are the modern drama of the te associated with Lessing
(birerlih Traurspil ) , with its Hellenic counterpart in Eurip
ides, representational music and the Florentine opera with its Hel
lenic counterpart in the Attic dithyamb. All are dismissed as
betrayals of the tragic origin of music. The difcult distinction be
tween the "bad" imitation of realistic art and the "good" imitation of
Wagner's music amounts, in fact , to the putting-into-question of the
identit between origin and end that shapes the genetic pattern of
the text itself. The condemnation of realism is frst carried ou
t
by
denouncing the overparticularization that reduces the godlike gener
ality of tragedy to the trva of imitative music, character dama,
or personal lyricism. But Nietzsche reaches beond the obvous
shortcomings of this conventional vew, and moves instead, with
sure hermeneutic instinct, to the sensitive points ofthe imitation, the
bnnin and the en of the work, the points where the validit of
the genetic pattern is at stak. In Eurpides, or in his moern equiva
lent, we are told that the authority, the reliabili t of the action in
terms of truth and falsehood that was implicitly given in the Sopho
clean and Aeschylean tragedy now has to be made explicit : "the
pathos of the exposition was lost [for Euripides] . Therefore, he put
the prologue even before the exposition and had it spoken b a
person one could trust . Ofen, some deit had to garantee the credi
bility of the stor and to remove all doubt about the reality of the
mythological plot , just as Descartes could only demonstrate the real
it of the empirical world by appaling to the truthfulness of a G
incapable of telling a lie" (3:88-89; 12) . The origin and beginning of
the narative is a literal , factual act of divine revelation and author
it. The same is true of the end: the same god must literally reappear
on the stage in order to resolve the apprently hopeless complications
and confer dignit upon the confusion of human enterprise. The
same "humanistic" pattern clearly applies to Th Birh ofTaedy, a
text based on the authorit of a human voice that rceives this au
thorit from its allegiance to a quasi-divne fgure. The prologue, an
invocation to Rchard Wagner, names the epiphany and vouchsafes
for the truth of the narrative, because Nietzsche "communicated
[with Wager] as if he were present and could therefore wite dow
only
t
hings worthy of this presence" (preface, 3: 19) . The resolution of
the narrative hinges on the rbirth of this same Spirit as the du e
96 RHETORI C
mahina, Wagner, reappearing on the scene to reverse the decadence
of art and lead the essay to its triumphant conclusion. To the extent
that it represents the histor of tragedy as the narrative of a sequen
tial event framed by the appearance and reappearance of the same
incarnate spirit , Th Birh a/Taey, as text , resembles a Florentine
opera or a birerlih Taurspil and not a Sophoclean tragedy or a
Wagnerian opra. It is therefore open to criticism directed against
these art forms. An intra-textual strcture within the larger strcture
of the complete text undermines the authority ofthe voice that asserts
the reliabilit of the representational pattern on which the text is
based. And it weakens the fgre precisely at the points that establish
its genetic consistenc: by weakening the authority of the pwer that
sustains, by its presence, the unit between the begnning and the end.
No wonder therefore that we must react with suspicion when
the discredited concept of representation is reintroduced in order to
distingish music from a purely imitative realism. Nietzsche cannot
give up the necessit for a representational moment as a constitutive
element of music. The tragic Dionysian insight is not , as for Rous
seau, an absence of all meaning, but a meaning that we are unable
to face for psychological or moral reasons. What Nietzsche calls,
following Schopenhauer, the "Will" is still a subject, a consciousness
capable of knowing what it can and what it can not tolerate, capable
of knowing its own volition. The self-representing facult of the will
is a self-willed act ; in music, the will wills itself as rpresentation.
Schopenhauer's defnition of music as being the "unmediated image
of the will [unmitelbars Abbild d Willen]" rests in the power and
the authority of the will as subject . Nietzsche can therefore only write
from the point of vew, as it were, of the will. The authority of his
voice has to legitimize an act by means of which the aporia of an
unmediated representation, by itself a logical absurdit, would be
suspended [au}ehoben] .
"One might say," writes Lacoue-Labarthe, "that Th Birh of
Tagey is ultimately nothing but the ambigous commentar ofthis
single statement by Shopenhauer [that music is the unmediated
image of the Will] , never accepted without reserations, but also
never truly contested. "9 Given the way in which Th Birh a/Taey
is rhetorcally organizd, Shopenhauer's dictum could only be "truly
contested" b undermining the authort of the narrator from within
9. Lacoue-Labarthe, "L Dtour," p. 70.
GENESI S AND GENEALOGY ( NI ETZSCHE)
97
the dynamics of the text . The negative valoriztion of repre
sentational realism and of the private lyrical voice (in Section 5) has
precisely this efect on the narrator-orator of Th Birh ofTagey.
Nowhere is this more apparent than when the exemplar value
of Wagerian opera has to be demonstrated. A narrative mythologi
cal superstructure that awakens a feeling of pity and sypathy has
to create the proper balance of distance and identifcation to "rescue"
the audience from "the unmediated contemplation of the highest
world-idea," "the unchecked efusion of the unconscious wll"(3: 144;
21) . "The myh shelters us from the music" (3: 141 ; 21) , if the Diony
sian trth of music as "Ding an sich" is to be maintained and if
music is the origin and the end of all art , then "pure," nonepresenta
tional music has to be literally intolerable.
I must address myself only to those who have a direct fliation
wth music, for whom music is lik a maternal womb, and
whose relationship wth things is determined almost exclusively
by unconscious musical ties. To these authentic musicians, I put
the question if they could imagine a human being able to hear
the third act of Tt an Isol wthout the assistance of
word and image, as if it were a single, overhelming sym
phonic movement? Such a listener would expire, carried away
on the overexpanded wngs of his soul . Could a man whose ear
had perceived the world's ver heart chamber, who has heard
the roaring desire for existence as if it were a thundering river or
the gentlest of brooks pouring out into the veins of the world,
fail suddenly to break down? How could he endure to hear the
echo of innumerable shouts of joy and pain, coming from "the
wide spaces of the world's night" and reaching him wthin the
mierable gass vessel of the human indvidual? Woud not this
metaphysical bacchanal compel him to fee back to his primor
dial home? [3: 143; 21]
Who would dare admit, after such a passage, to not being one of the
happy few among the "authentic musicians"? The page could only
have been witten wi th conviction if Nietzsche's personal identifca
tion would make him into the Kng Mark of a triangular relationship.
It has all the trappings of the statement made in bad faith: parallel
rhetorical questions, an abundance of cliches, obvious catering to its
audience. The "deadly power of music is a myth that can not with-
9
RHETORI C
stand the ridicule of literal description, yet Nietzsche is compelled,
by the rhetorical mode of his text , to present it in the absurdit ofits
facticit. The narrative falls into two parts or, what amounts to the
same thing, it acquires two incompatible narrators. The narator
who argues against the subjectivit of the lyric and aginst repre
sentational realism destroys the credibilit of the other narrator, for
whom Dionysian insight is the tragic perception of original truth.
It cannot be claimed that one of the narrators is merely the
Apollonian mask of the other. If this were the case, he could not be
making the claims for truth that are constantly being made in the
name of Dionysian wsdom. The myth, in Wagnerian opra, is not
just the dreamlike illusion that makes no claim beyond that of its
own beaut. It demands to be takn seriously as the only way of
access to a substantial truth. It ofers a necessar shelter to a full
consciousness and is not , like the Apollonian dream, the conscious
ness of a nonconsciousness. The Dionysian myth can no longer be
described and valorized in terms of appearance and ilusion, as
when the "will to delude" is presented as residing in a transcendental
force that accomplishes its. ow designs without concern for the
subject's own intentions. The Apollonian is by itself neither true nor
false, since its horizon coincides with the awareness of its ow illus
or nature. It is illusion and not simulacrum, for it dos not pretend
to be what it is not . The categories oftruth and falsehood can only be
introduced by the Dionysian subject which, from that moment on,
stands itself under the aegis of this polarit. Empowered, but also
compelled, to decide on matters of truth and falsehood-as when it
allows itselfto refer to the Apollonian as a lie ("suferng is somehow
being lied away out of the traits ofnatu"[3: 1 13; l6D-it has to run
the risk of having to decree the loss of its own claim to truth.
Have we merely been saying that Th Birh ofTagey i s self
contradictor and that it hides its contradictions by means of "bad"
rhetoric? By no means; frst of all , the "deconstruction" of the Diony
sian authorit fnds its argments within the text itself, which can
then no longer be called simply blind or mystifed. Moreover, the
deconstrction does not occur between statements, as in a logical
refutation or in a dialectic, but happens instead between, on the one
hand, metalinguistic statements about the rhetorical nature of lan
gage and, on the other hand, a rhetorical praxis that puts these
statements into a question. The outcome ofthis interplay is not mere
negtion. Th Birh ofTaed does more than just retract its own
GENESI S AND GENEALOGY ( NI ETZSCHE) 9
assertions about the genetic structure of literar histor. It leaves a
residue of meaning that can, in its turn, be translated into statement,
although the authorit of this second statement can no longer b lie
that ofthe voice in the text when it is read naively. The nonauthorita
tive secondar statement that results from the reading will have to be
a statement about the limitations of textual authority.
This statement cannot be read as such out of the original text,
although it is sufciently prepard ther to come to the surface in the
form of residual areas of meaning that cannot be ftted within the
genetic totality. Certain formulations in Th Birh ofTagedy remain
enigmatic and cannot be integrated within the value-pattern of the
main argment. For instance: afer having been consistently distin
guished from each other by a qualitative diferential system fouded
on the polarit between illusion and nonillusion, the Dionysian, Apol
lonian, and Scratic modes are at least once, in what seems like
casual aside, diferentiated in a purely quantitative system, in term
of their distance, as illusion, from a literal meaning. This literal
meaning then has the purely structural fnction of a degree of zero
fguralit and dos not, as such, coincide with tragic Dionysian in
sight. Contrar to all earlier claims, the Dionysian is then called one
stage of delusion [Iluwnstuf] among others, "the metaphyscal
consolation that the eternal life flows unimpaired beneath the tur
moil of appearances" ( 3: 121 ; 18) . Or what are we to make of the
theor of dissonance that comes to the fore near the end of the text
and fnctions there as a dynamic and temporal principle that can no
longer be called genetic? The semantic dissonance of Th Birh of
Tagedy is precisely this residue of meaning that remains beyond the
reach of the text's own logic and compels the reader to enter into an
apparently endless process of deconstruction. This process is itself
called "an artistic game that the will, in the eternal plenitude of its
pleasure, plays with itself (3: 161 ; 2) , a formulation in which ever
word is ambivalent and enigmatic, since the will has ben discre
dited as a self, the pleasure shown to be lie, the fullness to be absence
of meaning, and the play the endless tension of a nonidentit, a
pattern of dissonance that contaminates the ver source of the wll,
the wll as source.
A detour outside the main text is needed, not in order to resolve
the eniga, but to locate it, in its turn, within the context of its
rhetoricit. The lateral material for Th Birh of Tragedy that
Nietzsche lef out of the main essay contains formulations that re-
10 RHETORI C
translate into some knd of statement the disjunction between the
semantic assertion and the rhetorical mode that occurs in the main
text . We can confne ourselves, in conclusion, to two such statements.
The "tre rebuttal" of Shopenhauer, that Lacoue-Labarthe fails to
fnd in The Birh ofTagey, i . e. , the contestation of the will as the
ontological categor by means of which beginning and end, origin
and purpose are united in one genetic pattern, is brought to its
explicit conclusion in tIle discarded fragments, whereas it was
merely acted out theatrically in the main text . The following state
ment occurs in preparator outlines for Th Birh ofTagedy:
Intelligence is justifed in a world of aims. But if it is tre that
our aims are only a sort of rumination of experiences in which
the actual agent remains hidden, then we are not entitled to
transfer purposeful systems of action [Hanln nah
Zwekvrtllungen] into the nature of things. This means that
there is no need to imagine intelligence as capable of repre
sentation. Intelligence can only exst in a world in which mis
taks occur, in which eror reig- world of consciousness. In
the realm of nature and of necessit, all teleological hypotheses
are absurd. Necessit means that there can only be one possibil
it. Wy then do we have to assume the presence of an intellect
in the realm of things?-And if the will cannot be conceived
wthout impling its representation, the "will" is not an
adequate expression for the core of nature either. [Musarion,
3:239]
The radical separation of orign from purpose ( Urprung from
Zweck) that is established here eliminates all poSSible claim at ge
netic totalization. Dionysos, as music or as language, must now
belong either to the teleological domain of the text and then he is
mere error and mystifcation, or he belongs to "nature" and then he
is forever and radically separated from any form of art , since no
bridge, as metaphor or as representation, can ever connect the
natural realm of essences with the textual realm offorms and values.
It had always been stated, also in the published text of The Birh of
Tragedy, that Dionysos was not identical wth the Wil l ; he never i, in
the full sense, an essence, but the possibilit of an essence to exist in
the guise of its represented appearance. "One should here distinguish
as sharply as pOSSible the notion of essence from that of appearance,
GENESI S AND GENEALOGY ( NI ETZSCHE) 101
for the ver natur of music excludes that it be the will . This would
eliminate it from the realm of art altogether, for the will is in essence
the nonaesthetic. But music appears as will" (3:49; 6) . Th Birh of
Tagey damatizes a variet of manners by means of which the
distinction btween essence and appearance can be bridged; what
we have called the genetic pattern is precisely the possibility of this
bridge, of this translation (Nietzsche speaks of "ibersetzen" and
"iberbricken"IO) performed in the metaphorical narrative b means
of which Dionysos can enter into a world of appearances and stil
somehow remain Dionysos. The imager of fliation indicates that
this essence is able to function as origin, and thus allows the text to
unfold its symbolic stor. The unpublished fragments, contem
poraneous with the main text , deny this ver possibilit and thus
reduce the entire Birh of Tagedy to being an extended rhetorical
fction devoid of authorit. "One could object that I myself have
declared that the 'Will' receives an increasingly adequate symbolic
expression in music. To this I reply, in a sentence that summarizes a
basic principle of aesthetics: th Will i th object ofmui, but not it
orgn " (Musron, 3: 34) . Tis sentence coud never have stod in the
fal version if Th Bir of Tagey had to suive as a text. It is
hermeneutically stisfng however that the statement forced upon
us by the deconstruction of the main text would reach us, formulated
by the same author who also produced this text .
The deconstruction of the genetic pattern in Th Birh ofTaedy is
not without consequences, not only within the special feld of
Nietzsche interpretation, but in that of historiography and semiolog
as well. The dependence of narrative, continuous texts, such as Th
Birh ofTaedy, on discontinuous, aphoristic formulations, as in the
fragments from which the last quotations were taken, turns out to be
a recurrent structural principle of Nietzsche's work from the start .
From a historiographical point of vew, it is instructive to see a
genetic narrative function as a step leading to insights that destroy
the claims on which the genetic continuity was founded, but that
10. The image of the bridge appears in the main text within the sme context
ofunresolvable paradox: "We may make the form [of the opratic representation] as
visible and animate as possible, and have it glow with inner light , it still remains a
mere appearance from which no bridge leads us back into tre reality, into the heart
of the world" (3: 146; 21).
10 RHETORI C
could not have been formulated if the fallac had not been allowed
to unfold. This may well tun out to b an exemplar moel in trng
to understand the aberrant interretation of Rmanticism that
shapes the genealog of ou present-day historical consciousness.
Moreover, bearing in mind the analog that operates, in Th Birh of
Tagedy, between genetic movements in histor and semiological
relationships in language, the rhetorically self-conscious reading puts
into question the authorit of metaphor as a paadigm of poetic
language. For if genetic models are only one instance of rhetorical
mystifcation among others, and if the relationship between the
fgual and the proper meaning of a metaphor is conceived, as in this
text , in genetic terms, then metaphor becomes a blind metonyy
and the entire set of values that fgures so prominently in Th Birh of
Tagedy- melocentric theor of language, the pan-tragic con
sciousness of the self, and the genetic vision of histor-re made to
appear hollow when the are exposed to the clarit of a new ironic
light.
5
Retorc of
Tro
p
es
(Nitzh)
I T MAY S EEM FAR- FETCHED TO CENTER A CONS I DERATI ON
of Nietzsche's relationship t o literature on his theor of rhetoric. Why
should one choose to consider what , by all evdence, appears to be an
eccentric and minor part of Nietzsche's enterprise as a way of access
to the complex question of his refection on literature and on the
specifcally literar aspcts of his own philosophical di scourse? An
abundance of other, less oblique approaches to the question may
appear preferable. The confguration of the earlier literar examples
explicitly mentioned b Nietzsche, a constellation that includes a wide
variet of writers ranging from Gthe, Shiler, and H6lderlin to
Emerson, Montaige, and Sterne could certainly yield interpretative
insights. Or one could consider Nietzsche's literar ofspring, which
is certainly even more extensive and informative than one suspects.
The repertor of the revealed or hidden presence of Nietzsche in the
main literar works of the tentieth centur still has to b com
pleted. It would reveal many surprses of value to an understanding
of our period and literature in general. For Nietzsche is obviously
one of those fgures like Plato, Augstine, Montaigne, or Russeau
whose work straddles the two activities of the human intellect that
are both the closest and the most impenetrable to each other
literature and philosophy.
Nevertheless , the apparently crooked byways of the neglected
and inconspicuous corner of the Nietzsche canon dealing with
rhetoric will take us quicker to our destination than the usual itiner
ar that starts out from studies of individual cases and progresses
from there to sythetic generalizations. That this area has been ne-
1. As one example among many, I was strck to fnd many more traces of
Nietzche in Poust than assumed, ofen in connection with Wa
g
er and with the
t heme of music in general.
103
10
RHETORI C
gIected or discarded as a possible mainroad to central problems in
the interinterpretation of Nietzsche is clear frm bibliographical ev
dence: one of the few book dealing with the subject , a recent Gr
man work by Joachim Gth entitled Nitzch und die Rtorik
(Tibingen, 1970) , starting out from a suggestion that goes back to
Ernst Robert Curtius, remains strictly confned to stylistic description
and never pretends to engage wider questions of interpretation.
That , on the other hand, the consideration of Nietzsche's theor of
rhetoric, however marginal it may be, ofers at least some promise, is
clear from the work of some recent French commentators such as
Philippe Lacoue-Labarthe, Bernard Pautrat , Sarah Kofman, and
others. 2 Writing under the infuence of a renewed interest , in France,
in the theor of language, their work is oriented towards the
philosophical implications of Nietzsche's concerns wth rhetoric
rather than towards the techniques of orator and persuasion that
are obviously present in his style. I do not plan to deal wth these
particular contributions which are still preparator and tentative at
best , but wll tr instead to indicate, in too broad and too hast an
outline, how the question of rhetoric can be brought to bear on some
of Nietzsche's texts, early as well as late.
It is well known that Nietzsche's explicit concern wth rhetoric
is confned to the notes for a semester course taught at the University
of Basel during the wnter semester of 1872-73, with no more than
two students present . Parts of these notes have been published in
Volume V of the Kroner-Musarion edition. Only with their complete
publication, presumably in the new Colli-Montinari edition, wll we
be able to judge if the former editors were justifed in their claim
that, afer the seventh paragraph, the interest of the notes no longer
warranted their publication. It is also well kown that Nietzsche's
course on rhetoric was not original and drew abundantly on the
textbooks that were current at the t ime in the academic study of
classical rhetoric, especially Richard Volkmann, Di Rtorik dr
Griechn und Rmr in systmtichr
U
brsicht ( 1872) , Gustav
Gerber's Die Sprah al Kunst ( 187) and, on the question of elo
quence, the work of Blass ( 1868) .
3
There is sufcient manipulation
2. Se Brnard Paut rat, Verion du slil; Figre et swm d Nitzh
( Pari s, 1971 ) ; Sarah Kofman, "Nietzsche et la metaphore," Potiu 5 ( 1 971) : 77-9;
Philippe Lacoue-Labarthe, "I detour," Potiu 5 ( 1971) : 53-76.
3. Se Friedrich Nietzsche, "Retorique et langage," texts t ranslated, pre
sented, and annotated b Philippe Lacoue-Labarthe and Jean-Luc Nant in Poiu
5 ( 1 971 ) : 10.
RHETORI C OF TROPES ( NI ETZSCHE) 105
of these sources and sufcient new mphases in Nietzsche's notes
to justif their consideration despite their mied origns. To claim,
however, that they are of more than local signifcance takes some
more elaboration. At frst sight there is little in these notes to singe
them out for special attention.
Two main points that can be deduced from the notes desere to
be stressed. Nietzsche moves the study of rhetorc away from tech
niques of eloquence and persuasion [Beredamkeit] by makng these
dependent on a prevous theor of fges of speech or tropes. The
notes contain explicit discussion of at least three tropes: metaphor,
metonyy, and syecdoche, and announce Nietzsche's intention to
follow this up w
i
th a taxonomy of tropes that would include cata
chresis , allegor, irony, metalepsis, etc. Eloquence and style are an
applied form derived from the theor of fges. Nietzsche writes:
''There is no diference between the correct rules of eloquence [R]
and the so-called rhetorical fges. Actually, all that is generally
called eloquence is fgural language. "
4
The dependence of eloquence on fgure is only a further conse
quence of a more fundamental obseration: tropes are not under
stood aesthetcally, as ornament , nor are the understood seman
tically as a fgurative meaning that derives from literal , proper
denomination. Rther, the reverse is the case. The trope is not a de
rived, marginal , or aberrant form of language but the linguistic
paradigm par excellence. The fguative structure is not one linguistic
mode among others but it characterizes language as such. A series of
successive elaborations show Nietzsche characteristically radicaliz
ing his remarks until the reach this conclusion:
It is not difcult to demonstrate that what is called "rhetorical ,"
as the devices of a conscious art , is present as a device of un con
scious art in language and its development . We can go so far as
to sy that rhetoric is an extension [Forbilung] of the devices
embedded in language at the clear light of reason. No such
thing as an unrhetorical , "natural" language exists that could be
used as a point of reference: language is itself the result of
purely rhetorical trick and devices. . . . Language is rhetoric,
for it only intends to convey a dra (opinion) , not an epitm
(truth) . . . . Tropes are not something that can be added or
subtracted from language at will ; the are its truest nature.
4. Fredrch Nietzshe, Gammlte Werk ( Munich: Musaion Verlag, 1922) ,
5: 300.
10 RHETORI C
There is no such thing as a proper meaning that can be com
municated only in certain paricular cases.
Although it may seem daringly paradoxcal, the statement has
afnities with similarly orented formulations in Grber's Di
Sprahe al Kunst. This is not so surprising if one bears in mind
Grber's own antecedents in Grmn Rmanticism, especially in
Friedrich Shlegel and Jean Paul Rchter; the relationship of
Nietzsche to his so-called Rmantic predecessors is still largely
obscured by our lack of understanding of Rmantic linguistic theor.
Yet, the straightforard afmation that the paadigmatic strcture
oflanguage is rhetorical rather than representational or exressive of
a referential , proper meaning is more categorical , in this relatively
early Nietzsche text , than in the predecessors from which it stems. It
marks a fll reversal of the established priorities which traditionally
root the authority of the language in its adequation to an extralin
guistic referent or meaning, rather than in the intralinguistic re
sources of fgures.
A passage such as this one could still be understood as a belated
echo of earlier speculations, long since overcome in the post-Kntian
and pst-Hegelian sytheses that hav put rhetoric back in its proper
place, or dismissed it as a form of the aesthetic decadence that
Nietzsche will be one of the frst to denounce in later, anti
Wagnerian and anti-Schopenhaueran wrtings. The question re
mains however whether some of the implications of the early specu
lations on rhetoric are carried out in later work. At frst sight , this
hardly seems to be the case. The rhetorical vocabular, still much i
evdence in the Philosphbuh (which dates from the fall of 187
and thus immediately precedes the couse on rhetoric) disppears
almost entirely from Humn all to Humn on. It seems as if
Nietzsche had turned away from the problems of language to ques
tions of the self and to the asserion of a philosophy rooted in the
unmediated sense of existential pathos which has been so prevalent
in the interprtation of his work.
The validity of this scheme can be put in question by examining
one single but tical passage from a later text . It dates from 18
and is part of the posthumous fragments known as Th Will t
Power. The passage is characteristic of many later Nietzshe texts
5. Ibid.
RHETORI C OF TROPES ( NI ETZSCHE)
107
and is not to be considered as an anomaly. I am not primarily
interested in its specifc "thesis" but rather in the manner in which
the argument is conducted.
The passage has to do with what Nietzsche calls the
phenomenalism of consciousness, the tendenc to describe mental
events such as recollection or emotion in terms derivd from the
experience of the phenomenal world: sense perception, the interpre
tation of spatial strctures, etc. Under the heading "phenomenalism
of the inner world," Nietzsche writes as follows:
The chrnologal rersal which makes the cause reach con
sciousness later than the efect .-We have seen how pain is
projected in a part of the body wthout having its orign there;
we have seen that the perceptions which one naively considers
as determined by the outside world are much rather deter
mined from the inside; that the actual impact of the outside
world is never a cncious one . . . The fragment of outside
world of which we are conscious is a correlative of the efect
that has reached us from outside and that is then projected, a
posterri, as its "cause" . . . 6
The argment starts out from a binar polarity of classical banality
in the histor of metaphysics: the opposition of subject to object
based on the spatial moel of an "inside" to an "outside" world. As
such, there is nothing unusual about the stress on the unliability,
the subjectivit of sense impressions. But the workng hyothesis of
polarity becomes soon itself the target of the analysis. This occurs,
frst of all , b showing that the priority status of the to poles can be
reversed. The outer, objective event in the world was supposed to
determine the inner, conscious event as cause determines efect . It
tus out however that what was assumed to be the objective, exter
nal cause is itself the result of an internal efect . What had been
considered to be a cause, is, in fact , the efect of an efect , and what
had ben considered to b an efect can in its tun seem to fnction
as the cause of its own cause.
The to sets of polarities, inside/outside and cause/efect ,
which seemed to make up a closed and coherent sstem (outside
6. Fedrich Nietzhe, Wek in dri Bir, e. Karl Shlechta (Munich:
Hansr Verlag, 1956) , 3:80-05.
10 RHETORI C
causes proucing inside efects) has now been scrambled into an
arbitrar, open system in which the attributes of causalit and of
location can be deceptively exchanged, substituted for each other at
will . A a consequence, our confdence in the orignal , binar model
that was used as a starting point is bound to b shaken. The main
impact of this deconstruction ofthe classical cause/efect , subject/ob
ject scheme becomes clear in the second part of the passage. It is
based, as we saw, on an inversion or reversal of attributes which, in
this particular case, is sid to b temporal in nature. Logical priorit
is uncritically deduced from a contingent temporal priorit: we pair
the polarities outside/inside with cause/efect on the basis of a tem
poral polarity before/afer (or early/late) that remains un-refected.
The result is cumulative error, "the consequence of all previous
causal fctions," which as far as the "objective" world is concerned,
are forever tied to "the old error of original Cause. "
7
This entire
process of substitution and reversal is conceived by Nietzsche-and
this is the main point for us in this context-as a linguistic event. The
passage concludes as follows:
The whole notion of an "inner experience" enters ou conscious
ness only after it has found a language that the individual
unrstn-i .e. , a translation of a situation into afamiliar
situation-: 'to understand,' naIvely put merely means: to be
able to express something old and familiar.
8
Wat is here called "langage" is the medium within which the play
of rversls and sub
s
tittions that the passge describes takes place.
This medium, or propert of langage, is therfore the possibilit
of substituting binar plaritie such as bfor for afer, early
for late, outside for inside, cause for efect , wthout regard for the
truth-value of these strctures. But this is precisely how Nietzsche
also defnes the rhetorical fgre, the paradigm of all language. In the
Course on Retoric, metonymy is characterized as what rhetoricians
also call metalepsis, "the exchange or substitution of cause and ef
fect" and one of the examples given is, revealingly enough, the sub
stitution of "tonge" for language. Later in the same notes
metonymy is also defned as hypallags and characterized as fol
lows:
7. Ibid. , 3:85.
8. Ibid.
RHETORI C OF TROPES ( NI ETZSCHE) 109
The abstract nouns are properties within and outside ourselves
that are being torn away from their supports and considered to
be autonomous entities. . . . Such concepts, which owe their
existence only to our feelings, ae posited as if they were the
inner essence of things: we attribute to events a cause which in
truth is only an efect. The abstractions create the illusion as if
th were the entity that causes the properties, whereas the
receive thei objective, iconic existence [bildlich Dain] ony
from us as a consequence of these ver properties.
9
Pactically the same text that, in 187, explicitly defnes metonymy as
the prototpe of all fgural langage, describes, in 188, a metaphysi
cal construct (the phenomenalism of consciousness) as susceptible of
being deconstrcted as soon as one is made aware of its lingistic,
rhetorcal structure. We are not here concered with the conse
quences of this critique of phenomenalism which is also, in many
respects, a prefgative crtique of what wll later become known a
phenomenolog. Readers of Th Will to Powr kow that this critique
by no means pretends to discard phenomenalism, but puts us on our
guard against the tendenc to hypostatize consciousness into an au
thoritative ontologcal categor. Ad the wil also recognize that the
patter of argment here directed against the concept of conscious
ness is the same patter that underlies the critique of the main
categories that make up traditional metaphysics: the concepts of
identity, of causality, of the object and the subject, of truth, etc. We
can legtmately assert therefore that the ke to Nietzshe's critique
ofmetaphysics-which has, perhaps misleadingly, been described as
a mere rersal of metaphysics or of Plato lies in the rhetorical
model ofthe trope or, if one prefers to call it that way, in literature as
the language most explicitly grounded in rhetoric.
The idea of a reversal or an exchange of properties (in the
previous example, it is the exchange of the attributes of place and
causalit) is constittively paired by Nietzsche to the idea of eror:
the critical deconstruction shows that philosophical models such as
the phenomenalism of consciousness are indeed aberations whose
systematic recurrence extends throughout the entirety of classical
metaphysics. Would it not follow that, since the aberation turns out
to be based on a rhetorical substitution, it would sufce to bcome
9. Musaron, 5:319.
110 RHETORI C
awae of this i n order t o undo the patter and restore the properties
to thei "proper" place? If attributes of time and attributes of cause
have been improperly associated with each other, one migt be able
to uncross, so to speak, the polaities that have been exchanged in
order to recover a measue of truth. In the example at hand, we
could conceivably eliminate the misleading temporal scheme that
led to the confsion, and substitute for the derived cause, mistaenly
assumed to have an objective eistence in the outside world, an
authentic cause that could be infered from the critical deconstruc
tion of the aberrant one. Granted that the misinterpretation of realit
that Nietzsche fnds sstematically repeated througout the tradition
is indeed rooted in the rhetorical strcture of langage, can we then
not hope to escape from it by an equally sstematic cleansing of this
language from its dangerously seductive fgal properties? Is it not
possible to progress from the rhetorical langage of literatu to a
langage that, like the langage of science or mathematics, would be
epistemologically more reliable? The ambivalence of Nietzsche's at
titude towards science and literatue, as it appars, for example, in
the use of the term science in the title of "la gaya scienza" or in the
later fragments that look back upon Th Birth ofTagdy, indicates
the complexity of his position. One can read these texts as a glorifca
tion as well as a denunciation of literature. The general drif of
Nietzsche's thougt, on this point, can be better understood by ta
ing into account texts that precede the 1873 Course on Rhetoric,
especially the never-completed Philosophbuh.
For the ver question we are considering, the possibility of es
caping from the pitfalls of rhetoric by becoming aware of the
rhetorcit of language, is central to the entire Philoophnbuh and
its only completed unit, the essy On Tuth an Li in an Ea
Moral &ne [
U
ber Wahrhit un Luge im ausrmorlishn Sinn] .
This essay flatly states the necessar subversion oftruth by rhetoric as
the distinctive feature r f all langage. "Wat is truth?" ask
Nietzsche, and he answers:
A moving army of metaphores, metonymies and an
thropomorphisms, in short a summa of human relationships
that ae being poetically and rhetorically sublimated, trans
posed, and beautifed until , afer long and repeated use, a
people considers them as solid, canonical , and unavoidable.
Truths are illusions whose illusionar natur has been forgotten,
RHETORI C OF TROPES ( NI ETZSCHE) t:1
metaphores that have been used up and have lost their imprint
and that now operate as mere metal, no longer as coins.

What is being forgotten i n this false literalism is precisely the rhetori
cal , sybolic qualit of all language. The degradation of metaphor
into literal meaning is not condemned because it is the forgetting of a
truth but much rather because it forgets the un-truth, the lie that the
metaphor was in the frst place. It is a naive belief in the proper
meaning of the metaphor without awareness of the problematic
nature of its factual , referential foundation.
The frst step of the Nietzschean deconstruction therefore re
minds us, as in the above quotation, of the fguralit of all language.
In this text , contrar to what happens in Th Birh ofTagey, this
insight is openly stated as the main theme ofthe essay. Does it follow
that the text therefore escapes fom the knd of error it denounces?
Ad since we can mae the possibilit of this error distinctive of
literature in general , does it then follow that the essay On Li an
Tuth i s no longer literature but something closer to science-as
Wittgenstein's Tattu could claim to be scientifc rather than
literar? Or, if we call a hybrid text like this one "philosophical, " can
we then defne philosophy as the sstematic demytifcation of liter
ar rhetoric?
The text proceeds in its deconstructive enterprise by putting into
question some of the concepts that wll also be targets of the later
critique of metaphysics in Th Will t Power. It shows, for example,
that the idea of individuation, of the human subject as a privileged
viewoint , is a mere metaphor by means of which man protects
himself from his insignifcance by forcing his own interpretation of
the world upon the entire universe, substituting a human-centered
set of meaning that is reassuring to his vanit for a set of meanings
that reduces him to being a mere transitor accident in the cosmic
order. The metaphorical substitution is aberrant but no human self
could come into being wthout this error. Faced wth the truth of its
nonexistence, the self would be consumed as an insect is consumed
b the fame that attracts it . But the text that asserts this annihilation
of the self is not consumed, because it still sees itself as the center
that prouces the afrmation. The attributes of centralit and of
selfood are being exchanged in the medium of the language. Ma-
10. &hlechta, 3: 314.
112
RHETORI C
ing the language that denies the self into a center rescues the self
linguistically at the same time that it asserts its insigifcance, its
emptiness as a mere fgure of speech. It can only persist as self if it is
displaced into the text that denies it. The self which was at fst the
center of the langage as its empirical referent now becomes the
language of the center as fction, as metaphor of the self. Wat was
orignally a simply referential text now becomes the text of a text , the
fgure of a fgure. The deconstruction of the self as a metaphor does
not end in the rigorous separation of the two categories (self and
fgure) from each other but ends instead in an exchange of properties
that allows for their mutual persistance at the expense of literal
truth. This process is exactly the same as what Nietzsche describes as
the exemplar "lie" oflanguage: "The liar uses the valid designations,
words, to mae the unreal appar real . . . . He misuses the estab
l ished linguistic conventions by arbitrar substituton or en reer
sal of the names. "u By calling the suect a text , the text calls itself,
to some extent , a subject . The lie is raised to a new fgal power, but
it is nonetheless a lie. By asserting in the mode oftruth that the selfis
a lie, we have not escaped from deception. We have merely reversed
the usual scheme which derives truth from the convergence of self
and other by showing that the fction of such a convergence is used to
allow for the illusion of selfood to originate.
The pattern is perhaps clearest in the reversal of the categories
of good and evil as they combine with those of truth and lie. The
usual scheme derives good from truth and evl from falsehood. But
Nietzche tells the tale of the reversed pattern: in order to survive in
societ, man began by lying.
[Then] man forgets that this is the case: his lyng then is no
longer conscious and is founded on age-old habit-and it is by
thi nonawarnes, by this forgetting that he develops a sense of
truth. Because he feels obliged to designate a certain thing as
"red," another as "cold," a third as "mute," a moral impulse
oriented towards truth is awakened: in opposition to the liar,
who is trsted by no one and excluded from the group, man
discovers the respectabilit, the reliabilit and the use oftruth. 1
2
Thus moral
V
rtue is shown to originate out of lies. But the text
cannot go to rest in this deconstruction that would justit, to some
1 1 . Iid. , 3:31 1 ( my italics) .
12. Ibid. , 3:314.
RHETORI C OF TROPES ( NI ETZSCHE) 113
extent, the maralit af deceit (as we fnd it , far example, within a
palitical cantext , in Machiavelli ar in Russeau) . Far if we believe in
the maralit af deceit , we also have to believe in the evl af truth, and
to the extent that the saciet is held tagether by means af deceit, the
apn assertian af this fact will also destroy the maral arder. It cauld
hardly be said, withaut further qualifcatian, that a text like this ane
is sacially ar marally uplifing. Once again, the reversal af palarities
has nat led to a restaratian af literal trth-in this case, it wauld be
the assertian that maral educatian shauld increase ane's skll at
lying-but has drven u further into the camplicatians af rhetarica
delusian. We may have changed the rhetarical made but we certainly
have nat escaped from rhetaric. This cauld hardly have been ex
pected. The ariginal pairing afrhetaric wth error, as we encaunter it
fram the Caurse an Rhetaric to The Will to Pawer was based an the
crass-shaped reversal af praperties that rhetaricians call chiasmus.
And it turns aut that the ver process af decanstructian, as it
functians in this text , is ane mare such reversal that repeats the
selfsame rhetarical strcture. All rhetarcal structues, whether we
call them metaphar, metanymy, chiasmus, metalepsis, hypallagus ,
ar whatever, are based an substitutive reversals, and it seems un
likely that ane mare such reversal aver and abave the anes that have
already tan place wauld sufce to rest are things to their proper
arder. One mare "turn" ar trape added to a series af earlier reversals
wll nat stap the turn tawards errar. A text lik On Tth an Li,
althaugh it presents itselflegitimately as a demystifcatian afliterar
rhetaric remains entirely literar, rhetorical , and deceptive itself.
Daes this mean that it wll end up in a garifcatian af literature aver
science ar, as is sametimes claimed af Nietzsche, in a puely literar
canceptian af philasaphy?
Twa quatatians fram the Philaphnbuh, clasely cantempar
ar to On Tth an Li, fully reveal the ambiguit inherent in the
questian. On the ane hand, the truth-value af literature, albeit a
negative ane, is recagnized and asserted. Art is no langer assaciated
with the Dianysian immediac afmusic but is naw apenly Socratic in
its decanstructive functian. It is therefare, af all human activties, the
anly ane that can lay claim to truth: "Art treats appearance as ap
pearance; its aim is precisely not to deceive, it is therefare t. : ' 1 3 But
the truth af appearance, unlie the trth afbeing, is nat a threat ar a
passian that cauld be described in terms similar to thase used in Th
13. Musrion, 6: 9.
114 RHETORI C
Birth ofTagedy t o evoke the Dionysian pathos of trth. It can there
fore be said that it stands above pleasue and pai n, in the ordinar
sense of these terms. The artist, who is trthful in his recognition of
illusion and of lie for what they are, thus gains a special knd of
afective freedom, a euphoria which is that of a joYI wsdom or of
the Homerc Heitrkit and that difers entirely from the pleasue
principle tied to libido and desire. "As long as man look for truth in
th wrld, he stands under the dominance of desire [untr dr
Herrschaf d Tebe] : he wants pleasue, not trth; he wants the
belief in trth and the pleasurable efects of this belief. "14 Only the
artist who can conceive of the entire world as appearance is able to
consider it without desire: this leads to the feeling of liberation and
weightlessness that characterizes the man freed from the constraints
of referential truth, what Barthes, in more recent times, has referred
to as "la liberation du signifant. " On Tuth an Li describes the
euphoria of this tpe of "truth":
The intellect, this master of deceit , feels itself freed from its
habitual sertude when it is allowed to deceive wthout direct
harm. Then it celebrates its own saturnalia. It is never so rich,
so seductive, proud, clever and outrageous: with inventive satis
faction, it jugles metaphors and tears out [vnikt] the bor
dermarks of abstractions. For example, he considers the river as
if it were the moving roadway that carres man to where he
would otherwse have to walk. . . . It imitates human exis
tence as if it were a fne thing and declares itself entirely pleased
wth i t . 15
This attractive pairing of Heraclites with Stendhal is however not
devoid of warning signals . It has its ow psuedo-teleolog, the fow
of time delighting in the self-sufcient, innocent spectacle of its ow
motion. But if this movement is reduced to the mere appearance that
it is, it also loses its foundation and becomes one among the various
other metaphors of self-destrction disseminated throughout this
brief text : the insect and the futtering light , the conceptual pyramid
that turns out to be a tomb, the painter deprived of his hands, man
asleep on the back of a tiger. 16 The implicit threat in all these images
14. Ibid.
15. Shlechta, 3:320.
16. Ibid. , 3:310, 315, 317, 311, rsptively.
RHETORI C OF TROPES ( NI ETZSCHE) 115
is ver similar to the threat implied in mistakng a river for a road.
The crtical deconstrction that leads to the discover of the literar,
rhetorical nature of the philosophical claim to trth is genuine
enough and cannot be refuted: literature turns out to be the main
topic of philosophy and the model for the kind of truth to which it
aspires. But when literature seduces us wth the freedom ofits fgral
combinations, so much airier and lighter than the labored constructs
of concepts, it is not the less deceitful because it asserts its own
deceitful properties . The conclusion of the essay shows the artist in a
not particularly enviable situation: he is indeed freer but "he sufers
more [than the conceptual philosopher] whn he sufers; and he
sufers more ofen, because he does not learn from experience and
always again falls in the sme trap in which he fell in the frst place.
In his sufering, he is then just as foolish [unvminfg] as in his
happiness: he complains loudl and can fnd no consolation. "1 7 A
aphorism that dates from exactly the same period puts it more
bluntly and from a less personal point of view: it may be tre that art
sets the right norm for truth, but ''Trth kills, indeed kills itself
(insofar that it realizes its own foundation in error) . "" ! Philosophy
turns out to be an endless refection on its own destruction at the
hands of literature.
This endless reflection is itself a rhetorical mode, since it is
unable ever to escape from the rhetorical deceit it denounces. The
defnition of this mode lies beyond our present scope, though we get
some indication from the just-quoted description ofthe artist's plight
in On Tuth an Li as well as from the general tonality and struc
ture of this text . First of all, the description is certainly not a tragic
one: the sufering described in the passage, as well as the happiness
that precedes i t, cannot be taken seriously, since both are so clearly
the result of foolishness . The same foolishness extends to the text
itself, for the artist-author ofthe text , as artist , is just as vlnerable to
it as the artist-fgre described in the text. The wisdom of the text is
self-destructive (art is tre but truth kills itselO, but this self
destruction is infnitely displaced in a series of successive rhetorical
reversals which, by the endless repetition of the sme fgre, keep it
suspended between truth and the death of this truth. A threat of
immediate destruction, stating itself as a fgre of speech, thus be
comes the permanent repetition of this threat . Since this repetition is
17. Ibid. , 3:322.
18. Musron, 6:9.
116 RHETORI C
a temporal event , i t can be narrated sequentially, but what i t nar
rates, the subject matter of the stor, is itself a mere fgure. A non
referential , repetitive text narrates the stor of a literally destructive
but non tragic linguistic event . We could call this rhetorical mode,
which is that of the "conte philosophique" On Tth an Lie and, by
extension, of all philosophical discourse, an ironic allegor-but only
if we understand "irony" more in the sense of Friedrch Schlegel than
of Thomas Mann. The place where we might recover some of this
sense is in Nietzsche's own work, not in that of his assumed con
tinuators.
This conclusion as to the fundamentally ironic and allegorical
nature of Nietzsche' s discourse projects its efect on the work that
follow and on those that precede the Philosophenbuh as well as on
the relationship between the two segments that are thus being more
or less arbitrarily isolated. How an ironic readng of an allegorical
text such as Zarthutra or Th Genealog ofMorl, or the allegori
cal reading of ironic aphoristic sequences from The Gay Since or
Th Will to Powe would have to proceed cannot be outlined here,
however sketchily. I t may be more productive, i n conclusion, t o ob
sere how an early text such as The Birth ofTagedy fts into this
patter. For one of the most persistent ways in which the illusion that
rhetorical blindness can be overcome manifests itself is b the trans
ference of what Nietzsche calls "the old error of original cause" from
the statet to the hitor of the text. While granting the ambiva
lence of the later Nietzche on the subject of truth, one may contrast
this wariness wi th the relative naivete of the earlier work. Particular
texts from, say, On Tth an Li on, can be considered to be epis
temologically destructive, but b presenting them as a development
moving beyond the assumed mystifcation of the earlier writings, the
"histor" of Nietzsche's work as a whole remains that of a narrative
moving from false to true, from blindness to insight . But the question
remains whether the pattern of this narrative is "historical," i . e. ,
revelator of a teleologcal meaning, or "allegorical ," i. e. , repetitive of
a potential confusion between fgural and referential statement . Is
Nietzsche's work structured as a process, a movement of
"becoming"-and Nietzsche's late reference to "the innocence of
becoming" is well known-{r as a repetition? The importance of the
question is apparent from the near-obsessive way in which Nietzsche
himself, as well as his interpreters, have been returning to the enig
mas of the early Birth ofTagedy.
RHETORI C OF TROPES ( NI ETZSCHE)
117
The obvous pathos and exaltation of Th Birth afTagedy seems
entirely incompatible with irony. It is difcult not to read it as a plea
for the unmediated presence of the wll , for a trly tragic over an
ironic art . If this were indeed the case, then one would have to
assume a genuine development , even a conversion within Nietzsche's
thought during the years immediately followng the writing of Th
Birth afTagedy. The conversion could have been brought about by
his reflections on rhetorc as they appear in the Phlphbuh and
in the 1873 couse notes, and it would also be apparent in the
reaction against Wagner and Schopenhauer in the Unzeitgemie
Betrahtungen. The strcture of the work as a whole would then be
essentially diferent from that descrbed and acted out in On Tth
and Li.
A more rhetorically aware reading of Th Birth af Tagedy
shows that all the authoritative claims that it seems to make can b
undermined by means of statements provded by the text itself. And
if one also takes into account nates witten for Th Birth afTagedy
but not incororated in the published text , the ironization implicitly
present in the fnal version becomes quite explicit . Moreover, the
forthcoming publication, in the new critical edition of Nietzsche' s
work, of further lateral material for Th Birth afTagedy, shows that
the exclusion of these notes was dictated b considerations that dis
rpt the system of epistemological authority even more deeply. We
are tol d, in these fragments, that the valorization of Dionysos as the
primar source of trth is a tactical necessity rather than a substan
tial afrmation. Nietzsche's auditors have to be spoken to in Diony
sian terms because, unlike the Greek, they are unable to understand
the Apollonian langage of fgue and appearance. In pseudo
historical arguments, reminiscent of Holderlin's considerations on
the dialectical relationship between the Hellenic and the Western
world, Nietzsche wites: "The epic fable of the Ancients represented
the Dionysian in images. For us, it is the Dionysian that represents
(symbolizes) the image. In Antiquity, the Dionysian was explained
by the image. Now it is the image that is explained by Dionysos. We
have therefore an exactly reversed relationship. . . . For them, the
world of representation was clear; for us, it is the Dionysian world
that we understand. " 1 9 It follows that the entire system of valoriza-
19. Quoted from glley proofs of the forthcoming Colli and Montinari edition
of Th Birh ofTagey; no rference available.
118
RHETORI C
tion at work i n Th Birth ofTagey can be reversed at will. The
Dionysian vocabular is used only to mae the Apollonian mode that
deconstrcts it more intelligible to a mystifed audience. This ex
change of attributes involving the categories oftrth and appearance
deprives the to poles of their authorit. The binar polarit that
strctures the narrative of the text turns out to be the sme fgure we
have encountered in all previous examples, the same "reversal of
names" that was mentioned in On Tth an Li. If we read
Nietzsche with the rhetorical awareness provided by his own theor
of rhetoric we fnd that the general strcture of his work resembles
the endlessly repeated gesture of the artist "who does not learn from
experience and always again falls in the same trap. " What seems to
be most difcult to admit is that this allegor of errors is the ver
model of philosophical rigor.
6 Retorc of
Persuasion
(Nitzch)
THE QUESTI ON OF THE RELATI ONS HI P BETWEEN PHI L
osophical and literar discourse is linked, in Nietzsche, t o his critique
of the main concepts underlying Western metaphysics: the concept
of the one [hn] , the good [agathn] and the true [althia] . 1 This
critique is not conducted in the tone and b means of the arguments
usually associated with classical critical philosophy. It is ofen car
ried out by means of such pragmatic and demagogical value
oppositions as weakness and strength, disease and health, herd and
the "happy few," terms so arbitrarily valorized that it becomes
difcult to take them seriously. But since it is commonly admit ted
that value-seductions are tolerated (and even admired) in so-called
li terar texts in a manner that would not pass muster in "philosophi
cal" writings, the value of these values is itself linked to the possibilit
of distingishing philosophical from literar texts. This is also the
crudely empirical level on which one frst encounters the specifc
difcult of Nietzsche's work: the patent literariness of texts that
keep makng claims usually associated with philosophy rather than
wth literature. Nietzsche's work raises the perennial question of the
distinction between philosophy and literature by way of a decon
struction of the value of values.
The most fundamental ''value' ' of all, the principle of noncon
tradiction, gound of the identity principle, is the target ofa posthu
mous passage dating from the fall of 187:
We are unable to afrm and to deny one and the same thing:
this is a subjective empircal law, not the expression of any
"necessit" but only ofan inabilit.
1 . Eugen Fin, Nizh Philophi (Stuttgrt , 1960) .
119
120 RHETORI C
If, according t o Aistotle, the lw of contan i s the
most certain of all principles , if it is the ultimate ground upon
which ever demonstrative proof rests, if the principle of ever
axiom lies in i t ; then one should consider all the more rigorously
what presuppositions [Voraus etungen] already lie at the bot
tom of it. Either it asserts something about actual entities, as if
one already kew this from some other source; namely that
opposite attributes cannot be ascribd to them [konnen] . Or the
proposition means: opposite attributes shoul not be ascribed to
it [slln] . In that case, logic would be an imprative, not to
know the true [erkennn] but to posit [setzn] and arange a
world that shoul be tuefor u.
In short , the question remains opn: are the axoms oflogic
adequate to realit or are the a means and measure for us to
create the real , the concept of "reali t," for ourselves? . . To
afrm the former one would, as already stated, have to have a
previous knowledge of entities; which is certainly not the case.
The proposition therefore contains no criterin oftruth, but an
imperative concerning that which shoul count as true.
Supposing leetzt] there were no self-identical A, such as
is presupposed [vorugeetzt] by ever proposition oflogic (and
of mathematics) , and the A were already mere appearance,
then logic would have a merely apparent world as its precondi
tion [Vorusetzung] . In fact, we believe in this proposition
under the infuence of ceaseless experience which seems con
tinuously to confrm it. The "thing" -that is the real substratum
of A; our beliein things is the precondition [Voraus zung] of
our belief in logic. The A of logic is, like the atom, a reconstrc
tion [Nahkontruktin] of the "thing" . . . Since w do not
gasp this, but make oflogc a criterion of true being, we are on
the way to positing [setzn] as realities all those hyostases:
substance, attribute, object , subject , action, etc. ; that is, to con
ceivng c metaphysical world, that is a "true world" (his,
hower, is th apparnt worl one more . . . ) +
The ver frst acts ofthought , afrmation and denial , hold
ing true and not holding true, are, in as much as the presup
pose [vrusetzen] not only the habit of holding things true and
holding them not true, but the rigt to do so, already dominated
by the belief that thre i suh a thing a knowledge for u and
that jugnt really can rach th truth.in short , logic does
RHETORI C OF PERSUASI ON ( NI ETZSCHE) 121
not doubt its abilit to assert something about the true-in-itself
(namely that it can not have opposite attributes) .
Here reig" the coarse sensualistic preconception that sen
sations teach us truth about things-that I cannot say at the
same time of one and the same thing that it is hrd and that it is
sof. (The instinctive proof "I cannot have to opposite sens
tions at the same time"-uit carse andfa. )
The conceptual ban on contradictions proceeds from the
belief that we can form concepts, that the concept not only
desigates [behnn] the essence of a thing but compreh
it [as] . . . . In fact, log (lie geometr and arithmetic)
applies only to ftious truth (ngrt Wahrhitn] tht w
hav create. Logic is the attempt to unrstan th atual
worl by man ofa schm ofbeing positd [eetzt] by our
selve, mr correctly: to mak it easir to formalize and to
compute [berechnen] . . . . . 2
In this text, the polarities are no longer such spatial properties as
inside and outside, or categories such as cause and efect , or experi
ences such a pleasure and pain all of which fgre prominently in the
many sections in which consciousness or selfhood are the targets of
Nietzsche's critique. We are dealing wth the more elusive opposi
tions between possibilit and necessit, "k6nnen" and "sollen," and
especially between knowng and positing "erkennen" and "setzen. "
To kow [erknnn] i s a transitive function that assumes the prior
existence of an entit to be kown and that predicates the abilit of
knowng by ways of properties. It does not itself predicate these
attributes but receives them, so to speak, from the entit itself by
merely allowing it to be what it is. To the extent that it is verbal it is
prperly denominative and constative. It depends on a built-in con
tinuit within the system that unites the entit to its attributes, the
2. Friedrich Nietzsche, Werk Kth Geamtaugabe (KGW), ed. Giorgio
Colli and Mazzino Montinari ( Berli n: de Gryter, 1970) , 8(2) :53-5. In earlier edi
tions, the pssage appears as section 516 of Der Will zur Maht (for instance,
Musron edition, 19:23-29) . I quote in English from Fredrch Nietzsche, Th Will to
Power, translated by WaIter Kaufmann and R. J. Hollingdle (New York: Rndom
House, 1967) , pp. 279-80, wi th some slight modifcations for the ske ofterminolog
ical consistenc. All i talics are Nietzsche's. The sytactical form of the Grman terms
hs been al terd to avoid lenghty quotation. Se also note 9.
122
RHETORI C
grammar that lin the adjctive to the noun b predication. The
spcifcally verbal interention stems from the predication, but since
the predicate is nonpositional with regard to the properties, it cannot
b called a speech at. We could call it a speech fat or a fact that can
b spoken and, consequently, kow without necessarily introducing
deviations. Such a fact can, on the one hand, b spoken [knn]
without changing the order of things but it does not , on the other
hand, hve to be spoken [solln] since the order of things does not
depend on its predicative power for its existence. Knowledge [Er
knnti] depnds on this noncoercive possibility and in fact enun
ciates it by ways of the principle of the self-identit of entities, "the
self-identical A. "
On the other hand, language can also predicate entities: i n this
Nietzsche text , this is called "setzen" ( to posit) , the key verb around
which the logic of the passage tsts its snaelike way. ]t designates
genuine at of speech, the question being whether the identit
principle is an obligator speech act or a fact merely susceptible of
being spoken. Classical epistemolog, Nietzsche asserts, has main
tained the latter at least since Aristotle: " . . . according to Aristotle,
the law of contradiction is the most certain of all principles . . . , the
ultimate ground upon which ever demonstrative proof rests"; it is
the ground of all kowledge and can only b so by being a prr
given and not "put up," "gesetzt. " The deconstruction sets out to
show that this is not necessarily the case. The convincing power of
the identity principle is due to an analogical, metaphorical substitu
tion ofthe sensation ofthings for the kowledge of entities. A contin
gent property of entities (the fact that, as a "thing," the can be
accessible to the senses) is, as Nietzsche's early treatise on rhetoric
puts it , "torn away from its support"3 and fal sely identifed with the
entit as a whole. Li Rousseau, Nietzsche assimilates the delusive
"abstraction" of the "coarse snsualist preconception" with the possi
bility of conceptualization: the contingent, metonymic link of the
sensation [Empfndung] becomes the necessar, metaphorical link of
the concept : "The conceptual ban on contradiction proceeds from
the belief . . . that the concept not only desigates the essence of a
thing but comprehnd it. . . . " The semiological moment
[bezeihnen] , which can simply b described as the metonymic de
construction from necessit into contingenc, is clearly apparent in
this sentence. It asserts that , for Nietzsche as for Rousseau, concep-
3. Friedrich Nietzhe, Gammlt Werke CMinchen: Musrion; 1922), 5: 319.
RHETORI C OF PERSUASI ON ( NI ETZSCHE) 123
tualiztion is primarily a verbal process, a trope based on the sub
stitution of a semiotic for a substantial mode of reference, of sig
nifcation [bzihnn] for possession [asn] This is, however, only
one among a variet of deconstructive gestures and it is chosen for
strategc and historical rather than for intrinsic reasons.
For the text goes well beyond the assertion that the claim to
know is just an unwarranted totaliztion ofthe claim to perceive and
to feel . Elsewhere, Nietzsche will devote considerable energ to ques
tioning the epistemological authorit of prception and of
eudaemonic patterns of experience. But here he has other objectives.
The unwarranted substitution of knowledge for mere sensation be
comes paradigmatic for a wide set of aberrations all linked to the
positional power oflanguage in general, and allowng for the radical
possibility that all being, as the ground for entities, may be linguisti
cally "gesetzt ," a corrlative of spech acts. The text asserts this
without equivocation: "To afrm [that logical axoms are adequate
to reality] one would . . . have to have a previous knowledge of
entities: whih i crinly not th cae [my italics] . " It has, in truth,
not been shown explicitly that we have no a prr knowledge of the
being of entities. Wat has and wll be shown, within the confnes of
this particular fragment, is the possibility of unwarranted substitu
tions leading to ontological claims based on misinterpreted systems
of relationship (such as, for example, substituting identiy for sig
nifcation) . The possibilit of arousing such a suspicion sufcies to
put into question a postulate of logical adequac which might well
be based on a similar aberration. And since this aberration is not
necessarily intentional but grounded in the structure of rhetorical
tropes, it cannot be equated wth a consciousness, nor proven to be
right or wrong. It cannot be refuted, but we can be made aware of
the rhetorical substratum and of a subsequent possibility of error
that escapes our control . We cannot say that we know "das Seiende"
nor can it be said that we do not know it. Wat can be said is that we
do not know whether or not we know it because the knowledge we
once thougt we possessed has been shown to be open to suspicion;
our ontological confdence has forever been shaken.
Nietzsche seems to go further than this and concludes: "[The
law of contradiction] threore [my italics] contains no crtrn of
truth, but an impertive concerning that which should count as
true. " The conclusion seems irrevocable. As is stated at the beginning
ofthe passage (in the form ofa thesis) , the inability to contradict-to
state at the same time that A is and is not A-is not a necessity but
12
RHETORI C
an inadequac, "ein Nicht-vermogen. " Something one has failed to
do can become feasible again only in the mode of compulsion; the
perormative corelate of "I cannot" is "I [or you] must . " The lan
guage of identity and of logic asserts itself in the imperative mode
and thus recognizes its own activity as the positing of entities. Logic
consists of positional spech acts. A such, it acquires a temporal
dimension for it posits as futue what one is unable to do in the
present: all "setzen" is "voraussetzen," positional language is neces
sarily hypothetical . 4 But this hypothetical "voraussetzen" is in error,
for it presents a pre-positional statement as if it were established,
present knowledge. This blief can be deconstructed by showing that
the truths ofa logic based on noncontradiction are "fctitious trths. "
But in so doing the temporal order has also been reversed: it now
turns out that the future-projected, prospective assertion was in fact
determined by earlier assumptions, that the future truth was in fact
past error. All ''voraussetzen'' is "Nachkonstrtion" ( as when it is
said that the A of logic is "eine Nachkonstrution des Dinges") . The
deconstruction of the metaphor of knowledge into the metonymy of
sensation is a surface manifestation of a more inclusive deconstrc
tion that reveals a metaleptic reversal of the categories of anteriority
and posteriority, of "bfore" and "afer. " The "trth" of identity,
which was to become established in the future that follows its for
mulation turns out to have always already exsted as the past of its
aberrant "position. "
Does this mean that we can now rest secure ( thoug hardly
safe) in the knowledge that the principle of contradiction is aberrant
and that , consequently, all language is a speech act that has to be
performed in an imperative mode? Can we consequently free ou
selves once and forever from the constraints of identity by asserting
and denying the same proposition at the same time? Is language an
act , a "sollen" or a "tun," and now that we know that there is no
longer such an illusion as that of knowledge but only feigned trths,
can we replace knowledge by performance? The text seems to assert
this without question: it acts by denying the oneness and the same
ness of things. But in so doing it does not do what it claims to be
entitled to do. The text does not simultaneously afrm and deny
identity but it denies afrmation. 5 This is not the same as to assert
4. "man sollte elagen was (der Satz yom Widerspruch) im Grnde schon an
Behauptungen vraustzt. " KGW, 8(2) :53, II. 9-10.
5. Perhaps more dearly, in Grman: "Dr Text bejaht und verneint nicht ein
und dasselbe sondern er verneint daS Bejahen. "
RHETORI C OF PERSUASI ON ( NI ETZSCHE)
125
and to deny identit at the sme time. The text deconstructs the
authorit of the principle of contradiction by showing that this prin
ciple is an act , but when it acts out this act , it fails to perform the
deed to which the text owed its status as act .
The inconsistenc can be retraced by obserng the play of the
sme verb-root "setzen" in the following sentence: "Supposing
Ieetzt] there were no self-identical A, such as is presupposed [vr
auget ] by ever proposition of logc (and of mathematics) , and
the A were already mere appearn, then logic would have a merely
apparent world as its precondition [Voraustzun" The decon
struction oflogical and mathematical truth is based on the fact that
it is not rooted knowledge but that it depends on a prior act of
assumption [Voras zn] . This prior act is itself the target and the
outcome of the deconstruction. But the conclusion that would seem
to follow from this, namely that the principle of contradiction is to
be discarded, is agin formulated in a positional mode: "Getzt, es
gabe ein solches Sich-selbst-identisches A gar nicht . . . . " This ter
minolog is eminently correct, for we saw that the negtive proposi
tion ( there is no such thing as an A that is equal to A) has not been
established as kowledge (proven) but merely as a possibilit, a
suspicion-and any hypothetical knowledge is positional , Yet all
"setzen" has been discredited as unable to control the epistemologi
cal rigor of its own rhetoric, and this discredit now extends to the
denial of the principle of identit as well. The burden of proof shifs
incessantly back and forth between incompatible propositions such
as A A, A better be equal to A or else, or A cannot be equal to A,
etc. This complication is characteristic for all deconstructive dis
course: the deconstruction states the fallacy of reference in a neces
sarily referential mode. There is no escape from thi s, for the text also
establishes that deconstruction is not something we can decide to do
or not to do at will. It is co-extensive with any use of language, and
this use is compulsive or, as Nietzsche formulates i t, imperative.
Moreover, the reversal from denial to assertion implicit in decon
structive discouse neer reaches the symmetrical counterpart of
what it denies. In the sentence under discussion, for example, the
assertion that language is an act (the symmetrical counterpart ofthe
negative assertion that it is not a kowledge lased on the principle of
identit) cannot be taken as fnal: the term "gesetzt" functions as a
marker which undermines the authorit of such a conclusion. But it
does not follow that, ifit cannot be said of language that it is an act,
that it has to be a kowledge. The negative thrust of the deconstruc-
126 RHETORI C
tion remains unimpaied; afer Nietzsche (and, indeed, afer any
"text") , we can no longer hop ever "to know" in pace. Neither can
we exct "to do" anything, least of all to expurge "to know" and "to
do", as well as thei latent oppsition from ou vocabular.
Lest we b inclined to read this text as an ireversible passge
from a constative conception oflangage to a perfomative one, there
are several other statements from the sme general period in which
the pssibilit of "doing" is as manifestly being deconstructed as the
identit priciple, the gound of knowledge, is being put in question
here. This is not obvously the case: in many texts that are more
clearly destined for publication than the posthumous fragents, the
valoriztion consistently seems to privlege active forms of langage
over passive or merely reactive ones; the Genealg ofMorl is, of
couse, a clear case in point. Active and passive (or reactive) modes
are coordinated with values of hig and low or, more provocatively,
with those of master and slave, aristocrac and populace, distinction
and vlgrit. The passges from the Genealg on restimnt are
well know: rsentimnt is the state of mind of "such creatures that
are denied the true reaction, that of deeds . . . "; "In order to exist ,
slave moralit always fst needs a hostile external world; it needs,
physiologically speang, external stimuli in order to act at all-its
action is fundamentally reaction. The reverse is the case with the
noble mode ofvaluation. "6 And a little futher in the sme work, in
connection with a discussion of causalit that anticipates many simi
lar arguments in the posthumous fragments, the hystasis of action
as the horizon of all being seems to be unquestionably afrmed:
"there is no 'being' behind doing, efecting, becoming; the 'doer' is
merely a fction added to the deed-the deed is everthing" ["es gibt
kein "Sin" hinter dem Tun, Wirken, Werden; "der Tater" ist zum
Tun bloss hinzugedichtet-as Tun ist alles"7 The use of the term
"hinzugedichtet" (added b poetic invention) , as well as the context ,
indicate that action here is conceived in close connection with lin
guistic acts of wrting, reading and interpretation, and not within a
polait that opposes langage, as spech or as wrting, to action.
6. Murion, 15:295, quote in Engish from Friedch Nietzhe, O th
Gl ofMorl, ed. and trans. Walter Kufmann ( New York; Rndom Hous,
1
97) , fst essy, section 10, pp. 36-37.
7. Kumann, trans. , O th Gl ofMorl, frst esy, sction 13, p. 45;
Musrion, 15:305.
RHETORI C OF PERSUASI ON ( NI ETZSCHE)
:27
Of course, one cannot expect the same strateg wth regard to
valorization in a book like the Genealog explicitly designated as a
pamphlet and destined to condemn and to convnce, as in the more
spculative treatises that Nietzsche's later book (or books) were,
among other thing, destined to be. On a specifc question (such as
the ontologcal authorit of acts) the speculative statements should
be gven at least equal consideration next to the emphatic, persuasive
ones. One therefore has to confront a slogan such as "Tun ist alles"
wth a pssage like the followng: "The 'Spirt' , somthing tht thinks
here wefrst imagine an act that does not exist , 'thinking' , and
secon we imagine as substratum of this act a subject in which ever
act of thougt and nothing else originates: this means that th de
a wll a th doer are ftions [sowhl d Tn, al dr Tatr sin
fng] . "' The parallel that concerns us is the symmetr between
this fctitou doing fng Tn] and the fctitous trths fngr
Wahrhiten] that appar in the prevously discussed passge on the
principle of identity: "Logic (like geometr and arthmetic) applies
only to fctitious truths"
9
: here, in section 516, truth is opposed to
action as fction is oppsed to reality. In the later passage (section
477) , this conception of action as a "reality" opposed to the illusion of
knowledge is, in its turn, undermined. Perormative langage is
not less ambivalent in its referential function than the langage of
constatation.
It could be objected that, in the pssage now under discussion
( section 477) , it is not the realit of action in general that is being put
in question but specifcally the act ofthinkng and, furthermore, that
the linkge between the act and the performing subject ( the principle
of intentionality) is being deconstructed rather than action as such.
But Nietzsche is not concerned with the distinction between speech
(or thought) acts and, on the other hand, acts that would not b ver
bal . He is interested in the distinction between speech at and other
verbal functions that would not b perormative (such as kowng) . Non
verbal acts, if such a thing were to be conceivable, are of no concern
8. KGW, 8(2) :29, 1 1 . 9-17. Peviously published as sction 477 ofDer Will zur
Maht ( Musarion, 19:8) ; Kufmnn and Hollingdale, trans. , Th Will to Power,
p 263.
9. The earlier Nietzshe editions (i ncluding the Shlecta edition in 3 volumes)
all print "fngierte Wesenheiten" (fctitious entiti) but the Colli and Montinar
crtical edition gves "fngierte Wahrheiten" ( fctitious tuths) . The authortative ver
sion is more germane to our argument .
128
RHETORI C
t o him, since no act can ever be separated from the attempt at
understanding, from the interpretation, that necessrily accom
panies and falsifes it. The fctional truths, which are shown to be
acts, are always oriented towards an attempt "to unrstan the
actual world . . . to make it easier to formalize and to compute
[berchbar mahn] . . " and, in the later passage, thought is
also described as "an artifcial adjustment for the purpose of unr
standing [eine kintlich Zurchthung zum zwck dr
Verstindlichun]" ( 1296, ll. 8-9, my italics) . Even in the Genalog
the pure act that is said to b all there is, is conceived as verbal : its
paradig is denomination and the deconstruction of its genesis is
best caried out by means of etolog.
As for the intentional link between act and subject , it has been
the target of a considerable number of late texts, not to mention
several earlier versions that go back at least as far as the Birh of
Traged. In the posthumous texts, it is ofen caried out as a rhetori
cal deconstrction of the metalepsis of cause and efect ; the well
kow passage on the phenomenalism of consciousness is a good
case in point. 1o This moment in the deconstrctive proces is un
doubtedly still present in the fragment wth which we are concerned:
it is, afer all , entitled "On psycholog and epistemologl l and in it
Nietzsche denounces the acceptance of an "unmediated and causal
link between ideas" as "the crdest and clumsiest obseration. "12
There is nothing new about such utterances; what gives the passage
a special signifcance is that the fction ofa "subject-substratum" for
the act is explicitly called secondar as compared to the prior fction
of the act itself (''rst we imagine an act that does not exst . . . and
secon we imagine a subject-substratum for this act . . . ) . The aber
rant authorit of the subject is takn for granted; the new attack is
upon the more fundamental notion of "act . " Hence also the apparent
contradiction between this text and the one on the phenomenalism
of consciousness alluded to earlier ( Section 479) . Whereas the notions
of an "inner" space or time seem to be more or less defnitively
reduced to the status of a deception in the latter fragment, section
1 0. Sction 479 of Dr Wille ZUl Maht (Musarion, 19:10) ; Kufmann and
Hollingdale, trans. , Th Will t Power, 265-66.
1 1 . The heading appears only in the new Colli and Montinari critical edi tion,
8(2) :295.
12. KGW, 8(2) :295, II . 26-30; Kufmann and Holli ngdale, trans. , Th Will to
Power, p. 26.
RHETORI C OF PERSUASI ON ( NI ETZSCHE) 129
477 begins by asserting: "I maintain the phenomenalism of the inn
world, too . . . "; but the immediate continuation of the sentence
(" . . . everthing of which we become conscious is arranged, sim
plifed, schematized, interpreted through and through . . . and is
perhaps purely imaginar) makes clear that phenomenalit is now
no l onger used as an authoritative term that has to be deconstrcted,
but as the name of a metaphysical concept considered to be aber
rant . Section 477 takes for granted the deconstruction of the
phenomenalism of consciousness and of the subject caied out in
section 479 and it moves on to the more advanced target of "denken"
as act . If Nietzsche's notes were to be reordered as a logical proges
sion (in itself a nigtmarish and absurd assigment) , fragment 477
in the old classifcation would have to come afer fragment 479.
The deconstrction of thought as act also has a diferent rhetor
ical structure from that of consciousness: it is not based on metalep
sis but on secdoche: " 'Thinng,' as epistemologists conceive of it
[antz] , simply does not occur: it is a quite arbitrar fction,
arived at by singing out one element from the prcess and el iminat
ing all the rest, an artifcial arrangement for the purpose of intelligi
bilit. " 1 4 Wereas the subject results from an unwarranted reversal
of cause and efect , the illusion ofthought as action is the result of an
equally illegtimate totalization from part to whole.
The rhetorical structure of the fgures concern us less here than
the outcome of their analysis: the text on the principle of identit
established the universalit of the linguistic model as speech act ,
albeit by voiding it of epistemological authority and by demonstrat
ing its inabilit to perform this ver act . But the later text , in its turn,
voids even this dubious assurance, for it puts in question not only
that language can act rightly, but that it can be said to act at all . The
frst passage (section 516) on identit showed that constative lan
guage is in fact performative, but the second passage ( section 477)
asserts that the possibilit for language to perorm is just as fctional
as the possibilit for language to assert . Since the analysis has been
carried out on passages representative of Nietzsche's deconstructive
procedure at i ts most advanced stage, it would follow that , in
13. KGW, 8(2) :295, II. 15-22, Kaufmann and Hollingdale, trans. , Th Will t
Powr, pp. 263-6.
14. KGW, 8(2) :296, ll . 4-8; Kaufmann and Hollingdale, trans. , Th Will t
Powr, p. 26.
130
RHETORI C
Nietzsche, the cri tique of metaphysics can be describd as the decon
struction of the illusion that the langage of trth (epitm) could
be replaced by a language of persuasion (da) . What seems to lead
to an established priorit of "setzen" over "erkennen," oflanguage as
action over language as truth, never quite reaches its mark. It under
or overshoots it and, in so doing, it reveals that the target which one
long since assumed to have been eliminated has merely been dis
placed. The epitm has hardly been restored intact to its former
glor, but it has not ben defnitively elimi nated either. The difer
entiation between performative and constative language (which
Nietzsche anticipates) is undecidable; the deconstruction leading
from the one model to the other is irreversible but it always remains
suspended, regardless of how ofen it is repeated.
This conclusion takes us back to the Course on Retoric, which
precedes the posthumous fragments by ffeen years. The course
starts out from a pragmatic distinction between rhetoric as a system
of tropes and rhetoric as having to do with the sklls of persuasion.
[Beredamkitl Nietzsche contemptuously dismisses the popular
meaning of rhetoric as eloquence and concentrates instead on the
complex and philosophically challenging epistemolog of the tropes.
The distinction is not actually accounted for but taken over empiri
cally from the histor of rhetoric. Privleging fgue over prsuasion is
a typically post-Rmantic gesture and Nietzsche's dependance on his
predecessors in the Grman Rmantic tradi tion, from Friedrich
Schlegel on down, has been well documented. 15 The question, how
ever, is eternally recurrent and coincides wth the term "rhetoric"
itself. Within the pedagogical model of the trvum, the place of
rhetoric, as well as its dignit, has always been ambivalent: on the
one hand, in Plato for example and again at crucial moments in the
histor of philosophy (Nietzsche bing one of them) , rhetoric be
comes the ground for the furthest-reaching dialectical speculations
conceivable to the mind; on the other hand, as it appears in
textbooks that have undergone little change from Quintillian to the
present, it is the humble and not-quite-respectable handmaiden of
the fraudulent grammar used in orator; Nietzsche himself begins
his course by pointing out this discrepanc and documenting it wth
examples taken from Plato and elsewhere. 1 6 Between the two
15. Se, for instance, Philippe Lacoue-Labarthe, "Le Dtou," in Potiu 5
( 1971 ) : 53-76.
16. Musarion, 5:29.
RHETORI C OF PERSUASI ON ( NI ETZSCHE) t3t
functions, the distance is so wide as to be nearly unbridgeable. Yet
the two modes manage to exist side by side where one would least
expect it. Nietzsche's philosophical contempt for orator fnds im
pressive confrmation in the rigor of his epistemolog, yet , as any
reader of Th Birh ofTagedy, Th Gna ofMorl, or of that
irrepressible orator Zarathustra kows, there hardly is a trick of the
oratorical trade which he is not wlling to exploit to the full . In a
sense, Nietzsche has earned a right to this inconsistency by the con
siderable labor of deconstruction that makes up the bulk of his more
analytical witings. For this deconstrction seems to end in a reasser
tion of the active performative function of language and it rehabili
tates persuasion as the fnal outcome of the deconstruction offgural
speech. This would allow for the reassuring conviction that it is
legitimate to do just about anyhing wth words, as long as we kow
that a rigorous mind, fully aware of the misleading power of tropes,
pulls the strings. But if it turns out that this same mind does not even
know whether it is doing or not doing something, then there are
considerable gounds for suspicion that it does not know wht it is
doing. Nietzsche's fnal insight may well concern rhetoric itself, the
discover that what is called "rhetoric" i s precisely the gap that
becomes apparent in the pedagogical and philosophical histor of
the term. Considered as persuasion, rhetoric is perormative but
when considered as a system of tropes, it deconstructs its ow per
formance. Rhetoric is a t in that it allows for two incompatible,
mutually self-destructive points of vew, and therefore puts an in
surmountable obstacle in the way of any reading or understanding.
The aporia between performative and constative language is merely
a version of the aporia between trope and persuasion that both
generates and paralyzes rhetoric and thus gives it the appearance of
a histor.
Ifthe critique of metaphysics is structured as an aporia between
performative and constative langage, this is the same as saying that
it is strctured as rhetoric. Ad since, if one wants to consere the
term "literature," one should not hesitate to assimilate it wth
rhetoric, then it would follow that the deconstrction of
metaphysics, or "philosophy," is an impossibilit to the precise extent
that i t i s "literar. " This by no means resolves the problem of the
relationship between literature and philosophy in Nietzsche, but it at
least establishes a somewhat more reliable point of "reference" from
which to ask the question.
Pa II
Rousseau
7 Meta
p
hor
(Scn Durse)
THE PLACE OF THE Dicours on th Orgn an th Foun
tion ofInqulit among Men ( 1754) in the canon of Rousseau's
works remains uncertain. The apparent dualit of Russeau's com
plete wri ti ngs, a whole that consists i n part of political theor, in
part of literature (fction and autobiography) , has inetably led to a
divsion oflabar among the interpreters, thus bringing to light latent
incompatibili ties between political scientists, cultural historians, and
literar critics. This specialization has ofen prevented the corect
understanding of the relations between the literar and the political
aspects of Rousseau's thought. As the overtly political piece of writing
that it undoubtedly is, the Son Diourse has primarily interested
historians and social scientists. ! It does not confront them wth the
same difculties asjuli, a book in which it is not easy to overlook the
literar dimensions entirely and where it takes some degee of bad
faith to reduce the text to "an intellectual experiment in the tech
niques and consequences of human engineering."2 Despite the pres
ence of at least one explicit passage on language in the Dicoure the
linguistic mediations can easily be ignored. The section on the origin
oflanguage3 is clearly a polemical digression wthout organic lins to
the main argument , and the Dicoure can be considered as a literal
model for a theor of histor and of soiet, that is, a model that
1 . The bibliography of studies wholly or in part devoted to the Son Di
course is immense and one would welcome an updated eat d rchrh on the text.
In his notes to the edition of the Scn Diur in the Pleiade Eition (Pari) :
Gillimar, 19) Jean Starobinsk givs sveral useful indications (see notes, pp.
1297, 1299, 1305, 1315, 1317, 1319, 1334, 1339, 1359, 1370, 1372, 1377) . Since then
(196) there have ben numerous addtions.
2. Lester Crocker,j. ] Rusu (New York 196) , p. 20
3. Diours, pp. 146-51 . Al page references are to the French edition of the
Dicoure: Dicours sur l'orgn et l fonmt d l'inalit, texte etabli et anote
par Jean Starobinski, in]. ]. Russeau, Oeuv cmplt, ed. Brard Gagnebin and
Marcel Ryond (Paris: Gillimard [BibJiotheque de la Pleiade] , 19), vol . 3.
135
136
ROUSSEAU
could be transposed tel qul from the text to the political or soial
situation that it represents or prefgures. Once this is assumed, the
Scon Dicours becomes highly vulnerable to a list of recurrent
objections that reappear with remarkable persistence in all Rousseau
studies and that any reader of the text wll feel compelled to make
himself.
It is by no means my intention to sugest that these objections
are unfounded or that the are inspired by a deliberate malice that
should b met with defensive counter-mace. The Rousseau interpret
er should avoid the danger of repeating the paranoid gestue of his
subject. The frst task is to diagnose what, if anything, is being
sstematically overlooked by other readers, prior to asking why this
particular area of Rousseau's thought possesses the curious privlege
of rendering itself invisible, as if it were wearing the ring of Gyges
referred to in the sixth Pomn. The literal reading that fails to
take into account the fgural dimensions of the language (despite the
fact that this particular text explicitly draws attention to these di
mensions) is not to be rejected as simply erroneous or malevolent , all
the more since, in the Scon Dicours, the political terminolog and
the political themes postulate the existence of an extra-textual refer
ent and raise the question of the text's relationship to this referent .
Nor can we assume that this relationship i s one of literal corre
spondence.
Consider, for instance, the status of what seems to be the ines
capable a priori of the text itself, what Rousseau calls the "state of
nature. " Ver few informed readers today would still maintain that
Russeau's state of nature is an empirical realit, present , past , or
future. 4 Most commentators would agree that , at least up to a point ,
the state of nature is a state "that no longer exists, that has perhaps
never exsted and that probably will never come into being . "
(3: 123) . It is a fction; but in stating this, the problem has merely
been displaced, for what then is the signifcance of this fction wth
regard to the empirical world? Granted that the authorit of the state
of nature, the hold it has over our present thought, is no longer that
of something that existed elsewhere or at other times and towards
4. For a recent statement to this efect , among many others, % Henri
Guhier, L miitatons mtaphysiue d]. ]. Ruseau (Pars, 1970) , p. 23. For a
clear forulation of the fctional character of the state of natue, see Herbert
Dieckann's edition of Diderot , Supplment au vyage d Bougainvlle (Gneva,
1
95) , pp. lii-xciv.
METAPHOR (SECOND DISCOURSE ) 137
which our relation can therefore be described in terms of nostalgia
and quest; granted that the mode of being of the state of nature and
the mode of being of the present , alienated state of man are perhaps
radically incompatible, wi th no road connecting the one to the
other-the question remains why this radical fction ("We must
begin by discarding all facts . . . " [3: 132]) continues to be indis
pensable for any understanding of the present, as if its shadow con
trolled once and forever the degee oflight allotted to us . It is a state
that we must "kow well" and of which "it is necessr to have a
correct understanding [ds ntn jute] in order to evaluate our
present condition" (3: 123) . What knd of epistemolog can hope to
"kow well" a radical state of fction? The Son Dicourse hardly
seems to provide a reliable answer. As a genetic narrative in which
the state of nature functions at the ver least as a point of departure
or as a point of reference5 (if no longer necessaily as a point of
arrival) , the Scon Dicours seems to contradict the radical rejec
tion of realit on which it bases its claim to free itself from the
constraints of facts. Russeau seems to want to have it both ways,
givng himself the freedom of the fabulator but , at the same time,
the authorit of the responsible historian. A degee of impatience on
the part of the historians is certainly justifed towards a man who, b
his own admission, escapes in speculative fantasies but who, on the
other hand, claims that in so doing "one sweeps away the dust and
the snds that cover the edifce [of human institutions] , one reveals
the solid foundations on which it is built and learns to consider them
with respect" (3: 127) . How can a pure fction and a narative involv
ing such concrete political realities as propert, contractual law, and
modes of government coalesce into a genetic histor that pretends to
lay bare the foundations of human societ?
It seems difcult to avoid a prognosis of inconsistenc, leading
to the separation between the theoretical , literar and the practical ,
political aspects of Rousseau's thougt. The literar facult which, in
the Secon Dicourse, invents the fction of a natural state of man
becomes an ideolog growing out of the repression of the political
facul t. A clear and concise statement of this recurrent critical in
terpretation of Rousseau-which goes back at least as far as
Schiller-an be found in a recent study of the Scil Contrat by the
5. On this point, see Starobinsk's preface, Dicour, 3:1vii. He refers primaily
to an article b Eric Weil, ' :. J. Russeau et sa politique," Crtue 56 Oanuar 1952) :
3-2.
13
ROUSSEAU
French soial philosopher Louis Althusser. He analyzes recurrent
shifs [dalge] in the key terms of Russeau's vocabular and
concludes that these shifs, or displacements, are
to be explicitly understod, once and forever, as the ver dis
placement that separates the consequences of theor from real
it, a displacement beteen two equally impossible prae [d
cale entre du pratiu ia1emnt imposib1s] . Since we
now have [in the text ofthe Scial Contat] reached the stage of
realit and since w can only keep going around in a circle
(ideolog-conomy-ideolog, etc. ) no fight remains possible
into the actual, real world [dn l realit mm] . End of the
displacement .
If no other displacement is available to us . . . only one
single, diferent road remains open: a tranerenc [trarrt] of
the impossible theoretical solution into the other of theor
[l'aute d l thri] , namely literature. The fctional triumph
of an admirable, unprecedented literar work. 6
If the political side of Russeau' s work is indeed a reductive
ideolog that results from a repression carried out by means ofliter
ar langage, then the theoretical interest of a text like the Scond
Dicourse is primarly psychological . Conversely, the political writ
ings can then themselves become a reliable way of access to the
problematics of the self in Rousseau. And here the Scond Dicoure
woud b particularly useful , not only because, unlike the Scial
Contrat, it explicitly involves the moment of transference into liter
ar fction, but precisely because, unlike the autobiographical writ
ings, it hides its self-obsessions behind a language of conceptual
generalit. Rousseau's ambivalence wth regard to such key notions
as propert, civil authorit, and even technolog7 could then sere as
6. Luis Althusser, "Su Ie Contrat Social (Les decalages)" in Cahir pour
l
nl)s, 8, L'imp d}. } Rusu (Pars, 1970) , pp. 5-4.
7. The ambivalence of Russeau's attitude towards proprty is one example:
on the one hand, he makes it sound as if propert were thef; on the other hand, law
i at times glorifed, in almost extravagant terms, as the defense of proprty (se, for
exmple, Dours sur l'enomi politu, in Ouve cmpl, 3:2-49). One is
tempted to interpret the inconsistenc psychologically by rferring to Russau's
lowly birth as a soial misft who both glorifes proprty as something he desire but
cannot posess, and povrt as a self-redeeming moral vtue. On civil authorit, m
139 ROUSSEAU
a model for an understanding of his pschological self-myti
fcations. In strictly textual tens, the problem comes dow to
the inconsistenc between the frst and the second part of the text .
Between the pure fction of the frst part , dealing wth theoretical
problems of man, nature, and methoolog, and the predominantly
historical and institutional language, used in the second part, there
would exist a gap, an unbridgeable "dalg," that Russeau, caught
in a false claim of authentic self-kowledge, would be least of all
able to perceive. The reading that follows puts this scheme into
question.
In the Scon Dicourse, the state of nature, though fctional, is not
static. Possibilities of change are built into its description as a syn
chronic stte of being. The potentially dynamic properties of natural
man are pit, "a principle anterior to reason [that] inspires a natural
reluctance to see any sensitive being, and especially our fellow-man,
sufer or perish" (3: 126) , and freedom: "Nature alone does ever
thing in the actions of animals whereas man partakes in his ow
actions in his qualit as free agent" (3: 141) . The concept of pit has
been denitively treated by Jacques Derrida. 1 We can therefore begin
with the concept of freedom.
The ambivalent nature of the concept of freedom in Russeau
has been noticed b several interpreters. To be free, for Russeau, is
by no means a tranquil and harmonious repose within the ordained
boundaries of the human specifcit, the reward for a Kntian, ra
tional sense oflimitations. tom the start, freedom appears as an act
ofthe wll ("the wl still speaks when Nature is silent" [3: 141]) pitted
against the ever-present obstacle of a limitation which it tries to
transgress. 9 It is a consequence, or another version, of the statement
at the beginning of the Son Diours, that the specifcit of man
the disrepanc btween, on the one hand, the gorifcation of the magstrate of
Gneva and of his L father in the Dae ofthe&on Diure (3: 117-18) , and
the caricate of the harass mgstrate in the text of the Diours propr (3: 192-
9) .
8. Jacques Drd, De l Grammt
l
(Pars, 1967) , pp. 25972.
9. A summarzd in the admirable title of Starobinsk'sjen:aqu Ru
su: L tanparn et I'obtal (Par, 1957) . Rusau's statement to the Polish
naton is well k: "L reps et la librte sont incompatiles: il faut opter"
(Consiratins sur l GuWmt d Pi, Oeve compl, 3:955) . This
aspect of Russeau's thougt is now generally recognized in contemprar stdies.
140
ROUSSEAU
forever escapes our grasp since "the more we study man . . . the less
we are in a position to kow him" ( 3: 123) . Any confnement within
the boundaries of an anthropological self-defnition is therefore felt
to be a restriction beond which man, as a being devoid of natural
specifcit, wll have to transgress. This wll to transgress, in a pre
Nietzschean passage, is held by Russeau to be the ver defnition of
the Spiri t: "the pwer to will or, rather, the power to choose, as well
as the feeling of this power is a purely spiritual act" (3: 142) . Ver
little distinguishes power to will, or willpower (puisane d vou'ir)
from "wll to power," since the power to choose is precisely the
pwer to transgress whatever in nature would entail the end of
human power.
The direct correlative offreedom thus conceived is mentioned in
the paragraph that follows immediately upon the defnition, al
though the transitor link is not explicitly stated: freedom is man's
will to change or what Russeau somewhat misleadingly calls "per
fectibilit. "l o The potential transgression that occurs whenever the
concepts of nature and of man are associated-in the Esay on th
Orign ofLngge all examples destined to illustrate the "natural"
language of man are acts of volencel l-transforms all human at
tributes from defnite, self-enclosed, and self-totalizing actions into
open structures: perception becomes imagination, natural needs [be
soin] become unfulfllable passions, sensations become an endless
quest for knowledge all of which deprive man forever of a central
identit ("the more one meditates . . . the greater the distance be
comes between our pure sensations and the simplest forms of
kowledge" [3: 14]) . In the same consistent pattern, the discover of
temporality coincides wth the acts of transgressive freedom: time
relates to space in the same way that imagination relates to percep
tion, need to passion, etc. The ver conception of a future is linked
10. Misleadingly, since "perfectibility" is just as rgessive as it is progssive.
Starobinski, in a lengthy fotnote ( 3: 1317) assrts that perfectibilit i a "neologsme
savant"; the concept if not the wor appeas in Fontenelle's Dision sur l ancin
et l mm which dates from 16. Fontenelle spaks of "Ie progres des choses."
11. SEsai sur l'orgn d Ln, texte reprduit d'aprs l'oition A. Belin
de 1817 ( Paris, Ie Graphe, supplement au No. 8 des Cahir pour l'Anlyse) , hence
forth referd to as E ay, p. 502. Russeau mentions the theatening gfs sent by the
king of the Shians to Kg Darius and espcially the Old Tetament sto! Oudges)
o
f
the Levite from Ephraim who sent the boy of his murdered wife, cut in twelve
pieces, to the Tribs ofIsrael to spur them on to revenge. The same theme is taken up
in the later sto! L lit d'Ephraim (1762) .
METAPHOR (SECOND DISCOURSE ) 141
wth the possibilit of a free imagnation; the soul ofthe still enslaved
prmitive man is "without any awareness of the future, however
close it may be. His projects are as narow as are his vews: they
hardy extend until the end of the da ( 3: 14) . Consciousness of
mortalit is similarly linked to the freedom that distingishes m
from the animal : "the kowledge and the fear of death is one of the
frst thing acquired by man as he moves away from the animal
condition" (3: 143) .
This existential notion of freedom is impressive enoug in itself.
It does not sufce, however, to mae the connection with the politi
cal parts of the Sco
r
Dicoure. It accounts for the ambivalent
valorization of all historical change, since any change wl always
have to put into question the value-sstem that made it possible: any
positive valorization as progess always also implies a regess, and
Rousseau's text scrpulously maintains this balace. 12 The impossi
bilit of reaching a rationally enligtened anthropolog also accounts
for the necessar leap into fction, since no past or present human
action can coincide wth or be under way towards the natue of man.
The question remains why the Sco
r
Diourse, in its second part ,
somehow manages to retun to the concrete realities of political life
in a vocabular that reintroduces normative evaluations-why, in
other words, the methodologcal paradox of the beginning (that the
ver attempt to know man makes this kowledge impossible) does
not prevent the text from fnally getting started, afer many hesita
tions: a preface preceding a frst part which is itself a methodological
introduction and which, in its tun, is again introduced by another
preface. What characteristic strctures offreedom and perfectibilit,
in Part I, lead us to understand the political strctures of Part II? And
where are we to fnd a structural description of perfectibilit in what
seems to be a self-enclosed genetic text in which perfectibilit simply
functions as the organizing theme?
The section on langage ( 3: 146-51) appears as a digession
destined to illustrate the impossibilit of passing from natue to
culture by natural means. It runs parallel to a similar development
that deals with the gowh of technolog. As such, it seres indeed a
secondar function that belongs with the polemical and not with the
systematic aspects of the Sco
r
Dicoure. Starobinsk rightly em-
12. Se Diours, pp. 142, 162, 187, 193, es
p
ially note i, pp. 207-0, and
pasim.
142 ROUSSEAU
phasizes that the passage i s written "less i n order t o formulate a
coherent theor on the origin of langge than to demonstrate the
difculties the question raises" (3: 1322, notes) . In fact, the entie
passage has the tone of a mock-argument directed against those who
explain the origin oflanguage by means of causal categories that are
themselves dependent on the genetic power of the origin for which
they are sup
p
osed to account.
13
The constant warning aginst the
mystifcation of adopting a privileged viewoint that is unable to
understand its ow genealog, a methodological theme that runs
throughout the Scon Dicours, also applies to the theor of lan
gage. But not selectively s
O
. The science of language is one of the
areas in which this tpe of fetishism (reducing histor to nature)
occurs, but it is not the only one. The same error prevails with regrd
to ethical judgment (Hobbes) or with regard to technolog. From
this point of view, the section on langage seems to have a primarily
critical function and it could not sere to illuminate the central
problem of the text-that of the epistemological authorit of the
normative second part .
The passage, however, contains its ow theor on the structure
of langage, albeit in a highly fragentar and oblique form. More
important still , Rousseau explicitly link languag to the notion of
perfectibilit, itself derived from the primal categories of freedom
and will . "Moreover," he writes, "generl ideas can only enter the
mind by means of words and our understanding can seize upon
them only by means of propositions. This is one of the reasons why
animals could never acquire such ideas, nor the perfectibilit that
depends on it" ("C'est une des raisons pourquoi les animau ne
sauraient se former de telles idees, ni jamais acquerir la perfectibilite
qui en depend" [3: 149]) . Perfectibili t evolves as language eolves,
moving from particular denomination to general ideas: an explicit
link is established between two distinct conceptual areas in the text ,
the frst pertaining to perfectibilit, freedom, and a series of general
concepts that are connected narratively and thematically but never
described in terms of their internal structures, the second pertaining
13. . . . dire que I Mere dicte a l'Enfant les mots . . . cel montre bien
comment on enseige des Langes da formees, mais cel n'a
pp
rend pint com
ment elles se forment" C
p
. 147) ; "si les hommes ont bsoin de I
p
arole
p
ou
a
pp
rendre a
p
enser, i1s ont eu bien plus bsoin encore de savoi
p
enser
p
ou trouver
l'art de la
p
arole" C
p
. 142) . The conclusions reached b substituting efect for
cause Cmetale
p
sis) .
METAPHOR (SECOND DISCOURSE) 143
to the strctural and epistemological properties of language. Besides,
freedom and perfectibilit are relay-stations on the itinerar by way
of which the Scond Dicourse can move from the methoological
langage of the frst to the political langage of the second part. The
sentence can therefore b interpreted to mean that the stem of
concepts at work in the political parts of the Scon Discourse are
structured like the linguistic model described in the digression O
language. This makes the passage a ke to an understanding of the
entire text. For nowhere else do we fnd as detailed a structural
analysis of the concepts involved in the subsequent narrative.
Yet the passage is avoided rather than stressed in most readings
of the Scond Discourse. In his notes to the Pleiade Edition, Jean
Starobinski seems to be clearly aware of some of its implications, but
he at once limits its impact by means of an argument that goes to the
center of the problem involved in the interpretation of this text.
Commenting on Russeau's sentence-"C'est une des raisons pour
quoi les animaux ne sauraient se former des idees generales, ni
jamais acquerir la perfectibilite qui en depend"-he writes: "Te
relative clause [qui en depend] has here a determinative and not an
explicative function. Russeau refers here to one particular kind of
perfectibility that depends on language. As for Perfectibilit in gen
eral, which Rousseau has told us to be an essential and primitive
property of man, it is not the result of language but much rather its
cause" (3:1327, notes). Since the French language does not distin
guish between "which" and "that," it is impossible to decide by
grammatical means alone whether the sentence should read: "ani
mals could never acquire perfectibility, since perfectibility depends
on language" or, as Starobinsk would have it, "animals could never
acquire the knd of perfectibility that depends on language." The
corect understanding of the passage depends on whether one ac
cepts the contention that the principle of genetic causalit introduced
by Starobinsk, in which chronological, logical, and ontological prior
ity coincide,14 is indeed the system at work in Rousseau's text. Can it
be said of prfectibilit that it is an "essential and primitive property
of man," Starobinski's phrasing rather than Rousseau's, who said
only that it was "une qualite tres specifque qui distingue [l'homme]
14. Se Oeuv complie, 3: 125, notes. Starobinsk writes: "Russeau has
rigorously followed [Aistotle's] method, by gving to the word origin [arch] a
meaning in which the logical antecedent necesrly entails a hitorial antecedent."
14
ROUS SEAU
(3: 142) ? Each of the terms is problematic and their combination, as
if they could b freely interchanged, is the most problematic of all .
Starobinski's phrasing not only assumes that the (temporally) primi
tive must also b the (ontological) essence, but that a propert of
what is presumably a substance ( man) can be an essence. Since
moreover the substance "man" is in this text a highly volatile concept
that behaves logically much more like a propert than like a sub
stance, the essence perfectibilit would then be the propert of a
propert. Russeau's main methoological point , his constant warn
ing against the danger of substituting cause for efectl5 reveals at
least a certain distrust of genetic continuities, for the substitution
becomes aberrant only if such a continuit is in doubt. This should
mae us war of accepting uncritically the common sense and ad
mirable prudence displayed in Starobinski's reading.
Even if read to mean that perectibilit, in the general sense in
which it is used when we frst encounter it in the Scon Discourse
(3: 142) , is lined to language, the statement does not at frst sight
seem to be so far-reaching as to justif its repression. Why then is it
being overlooked or avoided? How curious that , when a text ofers us
an opportunit to l ink a nonlinguistic historical concept such as
perfectibilit to language, we should refuse to follow the hint . Espe
cially curious in the case of a text whose intelligibili t hinges on the
existence or nonexistence of such a link between a "literar,"
language-oriented method of investigation and the practical results
to which the method is assumed to lead. Yet a critic of Starobinski's
intelligence and subtlet goes out of his way in order to avoid the
sigs that Russeau has put up and prefers the bland to the sugges
tive reading, although it requires an interpretative efort to do so. For
there is no trace to be found in Russeau's work of a particular,
linguistic perfectibilit that would be distinct from historical prfect
ibilit in general. In the Essay on th Orgn ofLangge, the perect
ibilit of language, which is in fact a degradation, evolves exactly as
the perfectibilit of societ evolves in the Son Dicourse. There
must be an unsuspected threat hidden in a sentence that one is so
anious to de-fuse.
Animals have no histor bcause they are unable to perorm the
specically linguistic act of conceptualization. But how does concep
tualization work, according to Rousseau? The text yields information
15. Se
p
rding note 13 and 14.
M. ETAPHOR (SECOND DISCOURSE) 145
on this point, thoug not in a simple and straightforward way. It
describes conceptualization as substituting one verbal utterance (at
the simplest level , a common noun) for another on the basis of a
resemblance that hides diferences which permitted the existence of
entities in the frst place. The natural world is a world of pure
contiguity: "all indivdual entities appear in isolation to the mind [of
primitive man] , 16 as they are in the picture of nature. If one oak tree
was called A, another was called B . . . " ( 3: 149) . Within this con
tiguity certain resemblances appear. By substituting for A and B the
word "tree" on the basis of certain properties that A and B have in
common, we invent an abstraction under which the irreductible
diferences that separate A from B are subsumed. The perception of
these resemblances is not , in itself, a conceptualization: in the case of
animal s, it leads to acts that satisf needs but that remain confned
to the limits of the particular action. "When a monkey goes without
hesitation from one nut to another, do we think that he has in mind
a general idea of this tpe offruit and that he compares his archetpe
to these two indivdual entities? Certainly not . . . " ( 3: 149) . Concep
tualization dos not proceed on the basis of mere perception: percep
tion and imagnation ( in the gise of memor) 17 interene in recog
nizing the existence of certain similarities-an act of which animals
are said to be as capable as men-but the actual process of concep
tualization is verbal : "It is necessar to state propositions and to
spea in order to have general ideas; for as soon as the imagination
stops, the mind can only proceed by means of discourse" ( 3: 150).
The description seems to remain within a binar system in
which animal and man, nature and culture, acts ( or things) and
words, particularit (or diference) and generality, concreteness and
abstraction stand in plar opposition to each other. Antitheses of this
knd allow for dialectical valorizations and althoug this passage of
the Scon Dicourse ( 3: 149-50) is relatively free of value judgments
(nothing is said about an innate superiority of natue over artifce or
of practical behavior over speculative abstraction) , it nevertheless
invtes value judgments on the part of the interpreter. The most
incisive evaluations of this and of similar passages are those which
16. Rousseau sys "des premiers Instituteurs," hich may sound crtic in
translation. The meaning refers to "prmit" men as the "frst" inventors who
instituted language.
17. " ... la ve d'une de ces noi rappelle a [I] a memoire [du singe] les
sensations qu'il a re<ues de l'autre . . . " (3:150).
146 ROUSSEAU
locate the tension within language itself by stressing that the implied
polarit exists within the strcture of the linguistic sign, in the dis
tinction established by Russeau between the denominative and the
conceptual function of langage. The text indeed distinguishes the
act of naming (tree A and tree B) which leads to the literal denomi
nation of the proper noun, from the act of conceptualiztion. Ad
conceptualization, conceived as an exchange or substitution of prop
erties on the basis of resemblance, corresponds exactly to the classi
cal defnition of metaphor as it appears in theories of rhetoric from
Aristotle to Roman Jakobson.18 The text would then, in a sense,
distinguish between, on the one hand, fgurative, connotative, and
metaphorical language and, on the other, denominative, referential,
and literal language, and it would oppose the two modes anti
thetically to each other. This allows for a valorization that privileges
one mode over the other. Since Rousseau asserts the temporal priorit
of the proper noun over the concept ("Each object receivd frst a
particular name . . . " [3: 149] ; "thefrst nouns could only have been
proper nouns" [3:150] , it would indeed follow, within the genetic
logic of the narrative, that he separates the literal from the meta
phorical forms of language and privileges the former over the
latter. This interpretation, nearly unanimously accepted in Rousseau
studies, is well summarized, with a helpful reference to Michel
Foucault, by a recent commentator:19 "The entire histor of Rous
seau's work, the passage from 'theor to 'literature,' is the transfer
ence of the need to name the world to the prior need of naming
oneself To name the world is to mae the representation of the
world coincide with the world itself; to name myself is to make the
representation that I have of the world coincide with the representa
tion that I convey to others."2o Rousseau's increasingly subjective and
18. The defnition from the Poi (1457 b) is well kow: "Metaphor is the
transfer [epiphor] to a thing of a name that designates another thing, a transfer
from the genus to the spcies or from the spcies to the genus or according to the
prnciple of analog." Jakobson defnes metaphor as substitution on the basi of
rsmblance.
19. Alain Grosrchard, "Gravte de Russeau" in Cahirs pur I'anal
y
se, 8.
20. Grosrichard, "Gravte de Russeau," p. 6. I gve a free translation that
attempts to explain the more elegnt but more elliptical French version: "Toute
l'histoire de l'ouve de Russau, Ie pasge de la 'theorie' a la 'litteratur,' c'est Ie
passge d'une exgence qui est de faire se recouvir la representation du monde et Ie
monde meme, bref de Ie nomme, a l'exigence prealable de faie coincider la re
presntation que j'en donne a la reresentation que j'en ai, bref de m nommr."
METAPHOR (SECOND DISCOURSE) 147
autobiographical discourse would then merely be the extension,
within the realm of the self, of the referential linguistic model that
governs his thought. The failure of this attempt to "name" the sub
ject, the discover that, in Grosrichard's words, "Ie sujet est l'innom
able
,
,21 undercuts the authority of Rousseau's ow langage. It also
relegtes him, wth Condillac and, generally speaking, wth all fol
lowers of Locke, to what Foucalt subversively calls "Ie discours
classique." As far as the Scon Dicourse is concerned, such an
interpretation would have to conclude that the text is truly incoher
ent, since it does not control the opposition beteen the conceptual
metaphor "state of nature" and the literal reality of civil society, an
opposition asserted in the Dicours itself. Moreover, b starting out
from the metaphor, the text reverses the priority of denomination
over connotation that it advocates. In texts explicitly centered on the
self, such as the Cnfesin or the Dilog, this incoherence would
at least be brougt into the open, whereas it is merely repressed in
the pseudo-conceptual language of the Scond Dicourse.
Before yielding to this ver persuasive scheme, we must return
to the particular passage in the Dicourse and to the coresponding
section in the Eay on th Orign of Lngge. 22 Does Russeau
indeed separate fgural from literal language and does he privilege
one te of discourse over the other? There is no simple answer to
this question, for whereas, in the Dicoure, it is said that "the frst
nouns could only have been proper nouns," the Esay states wth
21. Ibid., p. 6.
22. On the complex debate involving the chronological and thematic rlation
ship btween the Diur and the E, sej. Dnd, D l Grammtol, pp.
272-78. One can consider the Eas an expanded footnote to the Iour. A far
as this particular point is concerned (animals lcking prfectibilit because the lack
conceptual language), the phrasing in the E y rns entirely prallel to the phrasing
in the Diours. The parallel is clos enough to allow for an extension of the
Dicur to include the Esy, at least on this pricular pint. "Les animaux qui
parlent (les langues naturlles) les ont en naissnt: ils les ont tous, et partout l
Meme; ils n'en changent pint, ils n'y font ps Ie moindre progeso L langue de
convention n'appartent qu'a l'hone. Voila purquoi l'homme fait des progs, sit
en bien soit en mal, et purquoi les animau n'en font pas" (Es, p. 50). That
"langue de convention" has the sme meaning as conceptual lnguage is part of our
argument. Starobinsk is cerainy rght to sy that "there is no contradction b
tw n (this text] and the pasge from the E y . . . " (3:1327). On the combined
reading of the E y with the Diour seals, for a diverent vew, Michele Duchet
and Michel Lunay, "Sychronie et diachronie; l'E sur l'rgn d lng et Ie
second Diours" in R inttn d Piphi, 82 (197): 421-42.
148 ROUSSEAU
equal assurance that "man's frst language had t o be fgrative" and
that "fgural language predates li teral meaning" (E ay, p. 50) . And
when we tr to understand denomination in Russeau as, in
Foucault's words, "going through language until we reach the point
where words and things are tied together in their common es
sence,,,23 then we fnd that, in the Son Dicoure, denomination is
associated with diference rather than wth identit. A note in the
1782 edition adds to the description of denomination ("If one oak
were called A, another would be called B") the following remark:
"for the frst idea we derive from two things is that they are not the
same; it ofen takes a great deal oftime to obsere what they have in
common." We would then have to assume that an obserer, so keenly
aware of difference that he fails to notice the resemblance beteen
one oak tree and another would be unable to distinguish the difer
ence between the word a and the tree A, to the point of considering
them as united in some "common essence. " Another difcult: fol
lowing the traditional reading of Russeau as it is here represented
by Alain Grosrichard, we would want to seize upon the act of de
nomination in all the transparenc of its nonconceptual literalness .
We fnd instead that "the frst inventors [of words] were able to give
names only to the ideas the already possessed . . . " (3: 150) , a sen
tence in which the word "idea," despite all pre-Kantian empiricist
concreteness, denotes the presence of some degree of conceptualit
(or metaphor) from the stat , wthin the ver act of naming. We
kow, moreover, from the previous quotation, what this ''ie pre
mire" must be: it is the idea of diference ("the frst idea we derive
from two things . . . ") . But if all entities are the sme, namely
entities, to the extent that they difer from each other, then the
substitution of sameness for diference that characterizes, for Rus
seau, all conceptual language is built into the ver act of naming, the
"invention" of the proper noun. It is impossible to say whether de
nomination is literal or fgral : from the moment there is denomina
tion, the conceptual metaphor of entit as diference is implied, and
whenever there is metaphor, the literal denomination of a particular
entit is inevitable: "tr to trace for yourself the image of a tree in
general , you wll never succeed. In spite of yourself, you wll have to
see it as small or large, bare or leaf, light or dark" or "As soon as you
imagine [a triangle] in your mind, it will be one specifc triangle and
23. Michel Foucault,L Mot et l ch (Paris, 196). The passage is quoted
in Grosrichard, "Gravte de Russau," p. 6.
METAPHOR (SECOND DISCOURSE) 149
no other, and it would be impossible not to make its contou visible
and its surface colored" (3:150) . Are we forced to conclude that
Russeau's paradoxes are genuine contradictions, that he did not
kow, in the Diours, what he stated in the Es y, and vice vers?
Perhaps we should heed his admonition: "in order not to fnd me in
contradiction with myself, I should be allowed enough time to ex
plain myself (Es y, p. 521) .
In the third section of the Es y on th Orn o Lngge,
Russeau ofers us an "example" in the form of a narrative parable, a
brief allegor. It tells us how the proper name mn, which fgures so
prominently at the beginning of the Scon Diours,2
4
came into
being:
A primitive man [un Imm sauvage] , on meeting other men,
will frst have experienced fright. His fear will make him see
these men as larger and stronger than himself; he wll give them
the name gnt. Afer many expriences, he will discover that
the supposed giants are neither larger nor stronger than him
self, and that their stature did not correspond to the idea he had
originally lined to the word giant. He will then invent another
name that he has in common with them, such as, for example,
the word mn, and will retain the word giant for the false object
that impressed him while he was bing deluded. [Esay, p. 506]
This is a general and purely linguistic version of what Grosrichard
calls "se nomm, " in which the origin of inequality, in the most
literal sense of the term, is being described. The passage was possibly
inspired, as has been pointed out,25 by Condillac, except for the fact
that Rousseau refers to full-gow men and not to childen. The
diference is important, for the entire passage plays a complex game
wth qualitative and quantitative notions of similarity, equality, and
diference.
In this encounter with other men, the frst reaction of the primi
tive is said to be fear. The reaction is not obvious; it is certainly not
based on objective data, for Russeau maes it clear that the men are
2. "L plus utile et l moins avancee de toutes les connaissnces humaines
me parait etre celie de l'homme . . . " (3:122). On the question of "man" in Rous
sau, see especially Martin Rng, J. J. Rusu Lhre vm Meh (Gttingen,
1959).
25. Among others b Starobinsk, Ouv comple, 3:1323, n. 3
150
ROUSSEAU
supposed to be of equal size and strength. Neither is it the fear of a
single individual confronted with a multitude, since primitive men
are entirely devoid of the sense of numbers or of groups. The similar
it in size and in the obserable attrbutes of strength should, at frst
sight , act reassuringly and make the reaction less anxious than if the
man had en
c
ountered a bear or a lion. Yet Rousseau stresses fright ,
and Derrida is certainly right i n stating that the act of denomination
that follows-alling the other man a gant, a process that Rousseau
describes as a fgral use of langage- displaces the referential
meaning from an outward, vsible propert to an "inward" feeling.26
The coinage of the word "gant" simply means "I am afraid. " But
what is the reason for fear, if it is not due to obserable data? It can
only resul t from a fundamental feeling of distrst , the suspicion that ,
although the creature does not lok like a lion or a bear, it neverthe
less might act like one, outward appearances to the contrar. The
reassuringly familiar and similar outside might be a trap. Fear is the
result of a possible discrepanc beteen the outer and the inner
properties of entities. It can be show that, for Rousseau, all
passions-whether the be love, pity, anger, or even a borderline case
between passion and need such as fear-are characterized by such a
discrepanc; they are based not on the kowledge that such a difer
ence exists, but on the hyothesis that it might exst , a possibili t that
can never be proven or disproven by empirical or by analytical
means.27 A statement of distrust is neither true nor false: it is rather
in the nature of a prmanent hythesis.
The fact that Rousseau chose fear as an example to demonstrate
the priorit of metaphor over denomination complicates and en
riches the pattern to a considerable degee, for metaphor is precisely
the fgre that depends on a certain degree of correspondence be
tween "inside" and "outside" properties. The word "gant ," invented
by the frightened primitive to designate his fellow-man, is indeed a
metaphor in that it is based on a correspondence between inner
26. Jacques Drrida, D 1 Grammt1, p. 393.
27. The assrtion has to b proven by a general interprtation of "passions" in
the work of Roussau. To indicate the direction of the argument, the followng
quotation frm Juli is characteristic; recapitulating the histor of her pssion for
Sint-Preu,Julie is sid to write: ':e crs vi sur votre visge les traits d I'ame qu'i1
fallaH a la mienne. II me smbla que mes sns ne seraient que d'orgne a des
sentiments plus nobles; et j'aimai dans vous, moins ce que j'y voyais que ce que j
croyais sentir en moi-meme . . .
"
(Ouv cmpl, 2:340).
METAPHOR (SECOND DISCOURSE) 151
feelings of fear and outward properties of size. It may be objectively
false ( the other man is not in fact any taller) but it is subjectively
candid (he seems taller to the frightened subject ) . The statement
maybe in error, but it is not a lie. It "expresses" the inner experience
correctly. The metaphor is blind, not because it distorts objective
data, but because it presents as certain what is, in fact , a mere
possibility. The fear offalling is "true," for the potentially destructive
power of gravit is a verifable fact , but the fear of another man is
hyothetical; no one can trst a precipice, but it remains an open
question, for whoever is neither a paranoiac nor a fool , whether one
can trst one's fellow man. By calling him a "giant ," one freezes
hypothesis, or fction, into fact and makes fear, itself a fgral state of
suspended meaning, into a defnite, proper meaning devoid of alter
natives. The metaphor "giant ," used to connote man, has indeed a
proper meaning (fear) , but this meaning is not really proper: it refers
to a condition of permanent suspense between a literal world in
which appearance and nature coincide and a fgural world in which
this correspondence is no longer a prr posited. Metaphor is error
because it believes or feigns to believe in its own referential meaning.
This belief is legitimate only within the limits of a given text : the
metaphor that connotes Achilles' courage by calling him a lion is
correct within the textual tradition of the Ili because it refers to a
character in a fction whose function it is to live up to the referential
implication of the metaphor. As soon as one leaves the text it be
comes aberrant-if, for example, one calls one's son Achilles in the
hope that this will make him into a hero. Russeau's example of a
man encountering another man is textually ambiguous, as all situa
tions involvng categorical relationships between man and language
have to be. Wat happens in such an encounter is complex: the
empirical situation, which is open and hypothetical , is given a consis
tenc that can only exist in a text . This is done by means of a
metaphor (calling the other man a giant) , a substitutive fgure of
speech ("h is a giant" substituting for "[ am afraid") that changes a
referential situation suspended between fction and fact (the
hypothesis of fear) into a literal fact . Paradoxcally, the fgre
literalizes its referent and deprives it of its para-fgural status. The
fgure dis-fgures, that is, it makes fear, itself a para-fgural fction,
into a reality that is as inescapable as the realit of the original en
counter between the two men. Metaphor overlook the fctional ,
textual element i n the nature of the entit it connotes. It assumes a
15
ROUSSEAU
world in which intra- and extra-textual events, literal and fgral
forms oflanguage, can be distingished, a world in which the literal
and the figural are properties that can be isolated and, consequently,
exchanged and substituted for each other. This is an error, although
it can be said that no langage would be possible without this error.
The intricac of the situation is obviously tied to the choice of
the example. The interplay of diference and simHari t implied in the
encounter between two men is more complex than if the encounter
had ben between two potentially antithetical entities such as man
and woman, as is the case inJuli or parts of Emil, or man and
things, as is the case in the example of the &con Diours in which
a man is naming a tree instead of naming another man. It seems
pererse on Rousseau's part to choose an example based on a more
complex situation than that of the paradigm wth which he is deal
ing. Should we infer, with the traditional interpreters of Rousseau,
that the intersubjective, refective situation of self-encounter, as in
the specular self-fascination of Narcissus, is indeed for Rousseau the
paradigatic experience from which all other experiences are de
rived? We must remind ourselves that the element of refective simi
larit mirrored in the example of man's encounter with man is not
the representation of a paradigmatic empirical situation (as is the
case in Descartes's cogto or in any phenomenological reduction) but
the metaphorical illustration of a lingistic fact. The example does
not hav to do with the genetic process of the "birth" of language
(told later in the text) but wth the linguistic process of conceptuali
zation. The narrative mode of the passage is itself a metaphor that
should not mislead us into transposing a synchronic, linguistic struc
ture into a diachronic, historical event. And conceptualization, as the
passage of the &con Dicur on the naming of trees makes clear,
is an intralinguistic process, the invention of a fgural metaanguage
that shapes and articulates the infnitely fragmented and amorphous
language of pure denomination. To the extent that all language is
conceptual, it always already speak about language and not about
things. The sheer metonymic enumeration of things that Rousseau
describes in the Dicur ("if one oa was called A, and another was
called B . . .") is an entirely negative moment that does not describe
language as it is or used to be at its inception, but that dialectically
infers literal denomination as the negation of langage. Denomina
tion could never exist by itself although it is a constitutive part of all
lingistic events. All langage is language about denomination, that
METAPHOR (SECOND DISCOURSE) 153
is, a conceptual, fgural, metaphorical metalanguage. As such, it
partakes of the blindness of metaphor when metaphor literalizes its
referential indetermination into a specifc unit of meaning. This
statement abut the metalinguistic (or conceptual) nature of lan
guage is the equivalent of the earlier statement, directly derived from
Rousseau, according to which denomination has to postulate the
concept (or idea) of diference in order to come into being.
If allianguage is about language, then the paradigmatic linguis
tic model is that of an entit that confronts itself.28 It follows that the
exemplar situation described in the Esa
y
(man confronting man)
is the correct linguistic paradigm, whereas the situation of the Sc
on Dicours (man confronting a tree) is a dialectical derivation
from this paradigm that moves away from the linguistic model
towards problems of perception, consciousness, refection, and the
like. In a text that associates the specifcit of man with language
and, within language, wth the power of conceptualization, the
priorit belongs to the example from the Esa
y
. The statement of the
Dicours that "the frst nouns could only have been proper nouns" is
therefore a statement derived from the logically prior statement
"that the frst language had to be fgural." There is no contradiction
if one understands that Rousseau conceives of denomination as a
hidden, blinded fgure.
This is not yt the end of the parable. Actual language dos not
use the imaginar word "giant"29 but has invented the conceptual
term "man" in its stead. Conceptualization is a double process: it is
this complexit that allows for the successive narrative pattern of the
allegor. It consists frst of all of a wild, spntaneous metaphor
which is, to some degree, aberrant. This frst level of aberration is
2. The implication is that the self-refective moment of the cto, the self
refection of what Rilke calls "Ie Narcisse exhauce," is not an original evnt but itself
an allegorical (or metaphorcal) version of an intralinguistic structure, with all the
negtive epistemologcal consequences thi entails. Similarly, Rousseau's use of
"fear" as the paradigmatic passion (or need) that leads to fgural language is not to
be accounted for in psychologcal but in linguistic terms. "Fear" is exemplr b
cause it correspnds structurally to the rhetorical moel of the metaphor.
29. The actal word "gant," as we kow it from everday usage, presupposes
the word "man" and is not the metaphorcal fgure that Roussau, for lack of an
existing word, has to call "giant." Russeau's "giant" would b more like some
mythological monster; one could think of Gliath, or of Polyphemos (leaving aside
the temptation to devlop the implications of Odyseus's strateg in giving his name
to Polyphemos as no-man).
15
ROUSSEAU
however not intentional, because it does not involve the interests of
the subject in any way. Russeau's man stands to gain nothing from
inventing the word "giant." The distortion introduced by the term
results exclusively from a formal, rhetorical potential of the lan
guage. The same is not true at the second stage. The word "man" is
created, says Rousseau, "afer many experences, [when primitive
man] will have discovered that the suppsed giants are neither larger
nor stronger than himself . . . " (E ay, p. 506) .. The word "man" is
the result of a quantitative process of comparison based on mea
surement, and makng deliberate use of the categor of number in
order to reach a reassuring conclusion: if the other man's height is
numerically equal to my own, then he is no longer dangerous. The
conclusion is wishful and, of course, potentially in error-as Goliath
and Polyhemos, among others, were soon enough to discover. The
second level of aberration stems from the use of number as if it were
a literal propert of thing that truly belongs to them, when it is, in
fact, just one more conceptual metaphor devoid of objective validity
and subject to the distortions that constitute all metaphors. For Rous
seau, as for Nietzsche, number is par excellence the concept that
hides ontic diference under an illusion of identit. The idea of
number is just as derivative and suspect as the idea of man:
A primitive could consider his right and his lef leg separately,
or consider them together as one indivsible pair, without ever
thinkng of them as t [legs]. For the representational idea of
an object is one thing, but the numerical idea that determines it
is another. Still less was he able to count up to fve. Although he
could have noticed, in pressing his hands together, that the
fngers exactly corresponded, he did not in the least conceive of
their numerical equalit. . . . [3: 219, note 14]
The concept of man is thus doubly metaphorical: it frst consists
of the blind moment of passionate eror that leads to the word
"giant," then of the moment of deliberate error that uses number in
order to tame the original wild metaphor into harmlessness (it being
well understood that this numerical terminolog of "frst," "doubly,"
"original," etc., is itself metaphorical and is used only for the clarit
of exosition). Man invents the concept man by means of another
concept that is itself illusionar. The "second" metaphor, which
Rousseau equates with the literar, deliberate and rhetorical use of
METAPHOR (SECOND DISCOURSE) 155
the spontaneous fgure30 is no longer innocent: the invention of the
word man makes it possible for "men" to exist by establishing the
equalit within inequalit, the smeness within diference of civil
society, in which the suspended, potential truth of the original fear is
domesticated b the illusion of identity. The concept interprets the
metaphor of numerical sameness as if it were a statement of literal
fact. Without this literalization, there could be no society. The reader
of Rousseau must remember that this literalism is the deceitful mis
representation of an original blindness. Conceptual language, the
foundation of civl society, is also, it appears, a lie superimposed
upon an error. We can therefore hardly eApect the epistemolog of
the sciences of man to be straightforard.
The transition from the structure of conceptual language to
society is implicit in the example from the Esay descrbing the
genealog of the word "man." It becomes explicit when, at the be
ginning of the second part of the Discours, the 6rigin of society is
described in exactly parallel terms, this time no longer as a marginal
example but as the central statement of theScon Discourse, forging
the axis of the text in the coherent movement that extends from
freedom to perfectibilit, from perfectibility to langage, from lan
guage to man, and from man to political societ. Neither the discov
er of fre and technolog, nor the contigity of man's proximity to
man on earth account for the origin of society. Society originates wth
the quantitative comparison of conceptual relationships:
The repeated contacts between man and various entities, and
between the entities themselves, must necessarily engender in
the mind of man the perception of relationships. These relation
ships, which we express b words such as large, small, strong,
weak, fast, slow, fearful, bold, and other similar ideas, when
compared to man's needs, produced, almost without his being
aware of it, some knd of refection, or rather some form of
mechanical prudence that taught him to tae the precautions
most needed for his safety ... . The resemblances that time
allowed him to obsere [between his fellow men], the human
female and himself, made him infer Uuger d] those which he
30. "L'image illusoire oferte par ia passion H montrant la premiere, Ie lan
gage qui lui repondait fut aussi Ie premier invente; il devint ensuite metaphorique,
quand l'esprit eciaire, reconnaissant s premiere erreur, n'en employa les expressions
que dans l es memes passions qui l'avaient prouite" (Ey, p. 506).
156
ROUSSEAU
could not perceive. Noticing that all of them behaved i n the
same way that he would himself have behaved in similar cir
cumstances, he concluded that their way of thinking and feeling
was entirely in conformity with his ow. [3:166]
3
1
The passge describes precisely the same interplay between passion
(fear), measurement, and metaphor (inferring invisible properties
by analog with visible ones) as the parable from the Eay on th
Orgn of Lngge. In the lines that follow, the principle of con for
mity on which the concept of man and the possibilit of government
is founded is called "cette importante Verite' (3:166). We should
now realize that what Rousseau calls "truth" designates, reither the
adequation of language to reality, nor the essence of things shining
through the opacity of words, but rather the suspicion that human
specifcity may be rooted in linguistic deceit.
The consequences of this negative insight for Russeau's politi
cal theor are far-reaching. What the Diourse on Inqulit tells us,
and what the classical interpretation of Russeau has stubbornly
refused to hear, is that the political destiny of man is structured like
and derived from a linguistic model that exists independently of
nature and independently of the subject: it coincides with the blind
metaphorization called "passion," and this metaphorization is not
an intentional act. Contrar to what one might think, this enforces
the inevitably "political" nature or, more correctly, the "pliticalit
(since one could hardly speak of "nature" in this case) of all forms of
human language, and especially of rhetorically self-conscious or
literar language-though certainly not in the representational, psy
chological, or ethical sense in which the relationship btween litera
ture and politics is generally understood. If soiety and government
derive from a tension between man and his language, then they are
not natural (depending on a relationship between man and things),
nor ethical (depending on a relationship among men), nor theologi
cal, since language is not conceived as a transcendental principal but
as the possibility of contingent error. The political thus becomes a
31. The translation considerably simplifes the opning lines, quite obscure in
the French text: "[L]'application reiteree des etres divers a lui-meme, et les uns aux
autres, dut naturellement engendrer. ... " The immediately preceding paragraph
in the text makes clear that Russeau refers to the interplay between several physical
entities in technological inventions (such as the invention of the bow and arrow) or
between man and nature, as in the discover and conseration of fre.
METAPHOR (SECOND DISCOURSE)
157
burden for man rather than an opportunit, and this realiztion,
which can be stated in an infnit of sardonic and pathetic moes,
may well account for the recurrent reluctance to accept, or even to
notice, the lin between language and societ in the work of Rus
seau. Far from being a repression of the political, as Athusser
would have it, literature is condemned to being the trly plitical
mode of discourse. The relationship of this discourse to political
praxis cannot be descrbed in pschological or in pscholinguistic
terms, but rather in terms of the relationship, within the rhetorical
model, between the referential and the fgural semantic felds.
To deelop the implications of this conclusion would lead to a
detailed reading of the second part of the Dicourse on Inqulit in
conjunction wth the Scil Contrat, Julie, and Russeau's other
political writings. I have tried to emphasize the importance and the
complexit of the transition that leads up to such a reading. Only if
we are aware of the considerable ambivalence that budens a
theoretical discourse dealing wth man's relation to man-"un
homme [que parle] a des hommes ... de l'homme," as the Scon
Dicourse puts it (3:131)-an we begin to see how Russeau's theor
of literature and his theor of government could get translated into
practical terms. The introductor analysis allows for the schematic
formulation of some directives.
First of all, the passage from a language of fction to a language
oriented towards plitical praxs implies a transition from qualita
tive concepts such as needs, passions, man," power , etc., to quantita
tive concepts involvng numbers such as rich, poor, etc.32 The in
equalit referred to in the title of the Dicourse, and which must frst
be understoo as diference in the most general way possible, be
comes in the second part the inequality in the quantitative distribu
tion of property. The basis of political thought, in Russeau, is eco
nomic rather than ethical, as is clear from the lapidar statement
that opens the second part of the Disourse: "The frst man who,
after havng fenced in a plot of land went on to say 'this belong to
m' and found other men naive enough to believe him [az simples
pour l croir], was the true founder of civl societ (3:16). The
32. "Que les mots de for et de faibl sont equivoues (dans Ie ca ou l'on
explique l'origine de l soiete par l'union des faibles entre eu); que ... Ie sens de
ces termes est mieu rendu par cu de pauv et de rh, parce qu'en efet un
homme n'avait point, avant les lois, d'autre moyen d'assujettir ses egu qu'en
attaquant leur bien, ou leur faisant quelque part du sien" (3:179).
15
ROUSSEAU
passge from literal greed to the institutional, conceptual law pro
tecting the right to propert runs parallel to the transition from the
spontaneous to the conceptual metaphor.33 But the economic found
ation of political theor in Rousseau is not rooted in a theor of needs,
appetites, and interests that could lead to ethical principles of right
and wrong; it is the correlative of linguistic conceptualization and i
therefore neither materialistic, nor idealistic, nor merely dialectical
since language is deprived of representational as well as of transcen
dental authority.34 The complex relationship btween Rousseau's and
Marx's economic determinism could and should only be approached
from this point of view.35
Second, one sees why civil order and government are, in Rous
seau, such fragile and threatened constructions, since the are built
on the ver sands of error.36 "The vices that make social institutions
necessar also mak the abuse of these institutions inevtable . . . "
(3: 187) . This circular, self-destructive pattern of all civl institutions
mirrors the self-destructive epistemolog of conceptual language
when it demonstrates its inabillit to keep literal reference and
fgural connotation apart. The literalism that makes language possi
ble also makes the abuse of language inevitable. Hence the funda
mental ambivalence in the valorization of literal reference through
out the Scnd Dicourse. The "pure" fction of the state of nature
precedes, in principle, all valorization, yet nothing can be more de
structive than the inevtable transpsition of this fctional model to
the present, empirical world in which "the subjects have to be kept
33. Thus confrming the semantic validity of the word-play, in French, on
"sens prpr" and "prpri. "
34. This, of course, dos not mean that questions of virtue, of self, and of G
are not bing considered b Rousseau; the obvously are. What is at stake i not the
existence of an ethical, pschologcal, or theological discours but thei authority i
terms of trth or falsho.
35. Hints in this direction are present in the work of Lucien Sbag,Marxm
e strturlim (Paris, 19), whereas Athusser remains short of Engels' treatment
of Roussau in the Anti-Dirn (espcially Chapter 13 of Par I, "Dalectics: Nega
tion of the negation"). I am not informed on the state of Roussau studies outside
Wester Europe and the United State.
3. ". . . ren n'est prmanent que la misere qui reulte de toutes ce vcis
sitdes; quand [l]es sentiments et [l]es idees [de l'homme]
p
uraient s'e1ever jusu'a
l'amour de l'ordre et aux notions sublimes de l vertu, il lui serait impsible de faie
jamais une application sure de ss princips dans un etat de choss qui ne lui
laissrait discerer ni Ie bien ni Ie mal, ni l'honnete homme ni Ie mechant" (Du
Cntat soil, Iere version, Oeve complt, 3:22).
METAPHOR (SECOND DISCOURSE) 159
apart" (Esay, p. 542) and by which one reaches "the last stage of
inequalit and the extreme point that closes the circle and touches
agin upon our point of departure (namely the state of nature): this
is where all individuals again become equal because they are nothing
. . . " (3: 191).
Finally, the contratul pattern of civl government can only b
understood against the background of this permanent threat. The
social contract is by no means the expression of a transcendental law:
it is a complex and purely defensive verbal strateg by means of
which the literal world is given some of the consistenc of fction, an
intricate set of feints and ruses37 by means of which the moment is
temporarily delayed when fctional seductions will no longer be able
to resist transformation into literal acts. The conceptual language of
the social contract resembles the subtle interplay between fgral
and referential discourse in a novel. It has ofen been said that
Rousseau's noveljuli is also his best treatise on political science; it
should be added that Th Social Contrat is also his best novel. But
both depend on their common methodological preamble in the
theor of rhetoric that is the foundation of the Dicourse on th Orign
an th Founation ofIne
q
ulit Among Men.
37. The furthest-reaching of these rses being perhaps that of the legslatr
havng to pretend that he speak wth the voice of Gd in order to be heard. "Voila ce
qui for<a de tout temps les peres des nations de recourir a l'interention du del et
d'honorer les dieux de leur propre sagesse, afn que les puples soumis aux lois de
l'Etat comme a celles de l natu, et reconnaissnt Ie meme pouvoir dans la
formation de l'homme et dans celie de la dte, obeissnt avec liberte et portassent
doilement lejoug de l fe1icite publique." Te example of the tre legislator is Moses
and the passage concludes with a footnote reference to Machiavlli. O. J. Rousseau,
Ouve cmplt, 3:383).
8
Sel
(Plin)
THE SECOND DISCOURSE IS THE STORY OF "A MAN [SPEAK
ing] of man . . . to men" (Ouvre Compltes, 3: 131). The situation
postulates an utterance (a man speang . . .), a meaning (about
man), and a reading (to men), a threefold articulation wthin the
Singe act of denomination. This structure is paradigmatic for all
cognitive discourse: it always has to b about an entit such as "man"
in which the noun is a conceptual metaphor that replaces a delusive
play between identit and diference. The naively empirical formula
tion of this fact (which Rousseau avoids) claims that man, as a
species, is defned by the possession of language as an elective attri
bute that is proper to him. But we do not "possess" langage in the
sme way that we can be said to possess natural properties. It would
be just as proper or improper to say that "we" are a propert of
language as the reverse. The possibilit of this reversal is equivalent
to the statement that all discourse h to be referential but can never
signif its actual referent. It leads to the loss of knowledge that
Russeau deplores at the beginning of the Dicourse: ". . . the least
advanced of all hum. an knowledge would seem to b the science of
man" but "what is more cruel still .. . is that, in a sense, by study
ing man we have made ourselves unable to know him" (3: 122-23), a
sentence in which "studyng must be understood as any process
involvng defnitional language. The relationship btween man and
his discourse is so far from being the simple possession of a natural
attribute that it has to be called "cruel" whereas, in the case of
natural properties such as the senses, it is the absence of the facult
rather than its existence that would arouse pit.
The &cor Discourse also demonstrates obliquely why all de
nominative discourse has to be narrative. If the word "man" is a
conceptual fgure grafed on a blind metaphor, then the referential
status of the discourse about man is bound to be curiously ambiva
lent. It claims to refer to an entit (man), but this entit turns out to
b the substitution of a defnitional for what was only a hypothetical
t6
SELF (PYGMALION) 161
knowledge, an epistemological metaphor substituting certitude for
ignorance on the basis of an assumed resemblance between passion
and prception, fear and size, inside ad outside. The resulting dis
course is complex not just because it has a pluralit of perhaps
incompatible meanigs but because the semantic status of any of
these meanings can never b determined. It always points to the
meaning which the fgure, by its ver existence, decrees: man as an
entit with specifc properties. But the substitutive chain that link
the fgure to its assumedly proper meaning can always be broken,
since it is grounded in hyothetical inferences that cannot be verifed.
In the case of such concepts as "fear," "state of nature," "passion,"
"perfectibilit and ultimately "man," it is impossible to decide
whether the are referential names for extralinguistic entities or
mere phantoms of language. And it is equally impossible to let the
question remain in abeance, since the pressure towards meaning
and the pressure towards its undoing can never cancel each other
out. This assetr is sugested, in Rousseau, by the stress on pas
sion (in which the rferential element is suspended) over need (in
which the referential element is determined) as the proper afective
metaphor for language! and, among passions, for such passions as
fear that remain, per defnition, in an intolerably suspended state.
Language can ony be about something such as man (i.e., concep
tual), but in being about man, it can never kow whether it is about
anything at all including itself, since it is precisely the abouts, the
referentialit, that is in question. Russeau's anthropological dis
course, as it comes to deal wth questions of selfood, of kowledge,
of ethical and practical judgment, of religion and politics, will al
ways be the restatement of this initial complication in a variet of
versions that cnfer upon his work an apparance of consistenc.
The classical polaities that shape narative discourse, such as
the distinction between showng and telling (mimesis and diegesis),
or, in a more recent terminolog, between dicours and hitoire, ae
correlatives of the initial complexit of denomination. Just as it is
impossible to say whether discourse about man is referential or not,
it is impossible to decide whether it is mimetic or diegetic. The
diegetic possibilit implies the hypothetical existence of a narator, of
1. Sa sur l'rig d ln, in Cirs pur L'Anly 8. Chapter 2 (p.
505) is entitled "Que la premiere invention de la parole ne vent pas des besoins, mi
des passions."
16
ROUSSEAU
a man talkng about man. It also implies the necessit of an act by
which the question of the referential verifabilit is raised, and this
epistemological moment, which cannot be short-circuited, is readily
represented in the fgure of an audience or a reader. We re-fnd the
traditional space or stage for the scene of reading as the scene of
telling, the mimesis of a diegesis. But the necessar presence of the
moment of utterance and of the interpretative moment of under
standing has nothing to do with the empirical situation naivly rep
resented in this scene: the notions of audience and of narrator that
are part of any narative are only the misleading fguration of a
linguistic strcture. And just as the indeterminac of reference gen
erates the illusion of a subject, a narrator, and a reader, it also
generates the metaphor of temporalit. A narrative endlessly tells the
stor of its own denominational aberration and it can only repeat
this aberration on various levels of rhetorical complexit. Texts en
gender texts as a result of their necessaril aberrant semantic strc
ture; hence the fact that they consist of a series of repetitive reversals
that engenders the semblance of a temporal sequence. All the con
stitutive categories of narration are implied in the theor of language
as fguration that appears in the Esay on th Orin ofLnge and
that is enacted in the fctional histor of the Son Diourse. One
should remember that they are the unfolding and not the resolution
of the chaotic uncertainty which Rousseau calls "fear."
But does all this complication and cruelt not stem from an
intrinsic weakness of the Scon Dicourse? Other theoretical texts
such as the Scil Contrat or the Pfesion d fi are much more
impersonal in tone than the hyrid Son Diourse, with its mixture
of polemics, pathos, argument, fction, and, at times, personal con
fession. On the other hand, openly autobiogaphical texts, from the
Ltr to Malhrbe ( 1762) to the Reri ( 1776-77) , though cer
tainly not devoid of generalizing intentions, are explicitly rooted in a
strong sense of particular selfood (one remembers Rousseau's claim
at the beginning of the Conesions : "Si je ne vaux pas mieux, au
moins je suis autre"2) . The complications of the Scon Discourse are
unquestionably related to the overgeneral use of the word "man" as a
conceptual metaphor, detached from its empirical foundation in the
re
fl
ective self -experience of a subject, in the constitutive cogto of a
2. J. J. Russeau, Ouve complt, ed. Bernard Gagnebin and Marcel
Rymond ( Pais: Gallimard [Bibliotheque de la Pleiade] , 1959) , 1 :5.
SELF (PYGMALION) 16
consciousness. If one admits, wth Alain Grosrichard and many
others, that the main pupose of Rousseau's work is "to make the
representation I have of the world coincide wth the representation I
convey to others or, in brief, to name myself [m nommr] ,"3 then
this act of self-denomination remains confused, in the Scond Di
coure, wth the anthropological defni tion of man in general. The
fact that the mediations leading from the particular to the general
self remain repressed [impe] would then explain the extreme
ambivalene, bordering on incoherence, of the Dicourse. The prob
lem may never be resolved in Rousseau's further work but, in com
pensation, the near-obsessive concentration on a partly censored self
that becomes more and more evdent as one moves from the main
political writings to, say, the Dialog (1773) leads to a refnement
of self-insight that places Rousseau in the main tradition of the
post-Augustinian literature ofthe self. He restates the Delphic "kow
thyself' alluded to at the onset of the Scon Dicours in the
eighteenth-centur terminolog of the "morale sensitive. " Tat such
acute self-understanding is by no means incompatible with patho
logical misinterpretations of the selfs relationship to others and to
the world is by now a psychological commonplace. It is also well
known that the interpretation of Rousseau as a philosopher of the self
has been among the most productive ways to read him in past as
well as in recent times, from Hazlitt and Grmaine de Stael to Jean
Starobinsk.
Before turning to the further development of Rousseau' s fgural
rhetoric, we must therefore consider his ow understanding of self
hood. Do we indeed avoid some of the complications of the Scon
Diourse by starting out from the particular self instead of starting
out from the idea of man in general? Is the theor of metaphor
derived from the Scon Dicourse still applicable in texts centered on
self-refection rather than on historical fctions? More specifcally,
can the self be called a metaphor in the same way that "giant" and
subsequently "man" could be show to be the result of coercive
tropological displacements?
The word "metaphor" occurs rarely in Rousseau, who speak more
generally of "langage fgre' wthout making distinctions between
particular tropes or getting involved in the pitfalls and refnements of
3. Aain Grosrichard, "Gravte de Russeau" in Cahs pur I'any, 8, p. 6.
16
ROUSSEAU
such distinctions. I t does appear however i n a piece of dialoge
spoken by a ver minor character in a ver minor work, the early play
Narcis e that Rousseau claims to have witten at the age of eighteen.
It tells the stor of a character named Val ere so inebriated by vanit
that he falls in love with his ow portrait, barely disgised as a
woman. His valet Frontin, whose inebriation, afer makng the
rounds of the cit cabarets, is a great deal more literal, is in conversa
tion with Lucinde, Vaiere's sister:
Luin: He is drunk, I believe. Ah, Frontin, I beg you, tr to get hold
of yourself and to make some sense.
Fontn: Nothing could be easier. See here. It is a prtrait . . . a
portrait that has been metamor . . . no metaphor . . . yes,
metaphorized. It's my master, it's a girl . . . you have made a
certain mixture . . . I have guessed all that, I . . .
[Naris, 2: 100]
The exlicit association of "portrait"4 wth metaphor (leaving
aside the further complications introduced by "metamorphosis" and,
for that matter, by drunkenness) allows for a prallel with the fable
of the Es y. The passion involved in Narise is not, as in section II of
the Essay, "fear" but "love" or, more precisely, the interplay between
self-loe (amour d soO, vanit (amour propre) , and the love of
others that maes up the passion. Already in this early text (revsed at
a later moment that is hard to determine)5 love is constitutively
associated wth the notion of self The frst explicit formulation of the
structure of self-love, in note 15 to the Scn Dicourse, makes clear
that "loye" is characterized by the diferential relationship between
the prsonal and the reflexive pronoun in the paradigmatic sentence
that designates self-Iove:je m'aime. "In the true state of nature," sys
Russeau, "vanit [amour prpr] does not exist, for each particular
4. "Porrait" will become, of course, a particularly rch and ambigous term
throughout Russeau's work, from the ealy riddle ( 2: 1 133) , to the portrait inNai,
the sene oOulie's porrait in the Nouvl Heli (2:2788), the disticton bten
"po
r
rit" ad "tableau" in the Scond Pefce to the Nouvl Heli, and the ranting
abut his ow portrait in the Dig ( Dueme dialoge, 1 : 777 f. and notes) .
5. On the complex histor of Naris, se the notes oOacques Sherr in his
edition of Narie in J. J. Russeau, Ov complte, ed. Brnard Gagebin and
Marcel Ryond (Paris: Gallimar [Bibliotheque de la Pleiade] , 1961 ) , 2: 1858-65.
"On the
poSSible corrections made b Marivau," wites Sherer, "one can only make
conjectres" (2: 1860) .
SELF ( PYGMALION) 165
man considers himself as the only spctator that takes notice of him ,
as the sole being in the universe that takes an interest in his existence,
as the only judge of his own merit. It is therefore not possible for a
feeling to burgeon in his soul that takes its origin in comparisons
which he is not able to can out . . . " (Son Dicours, 3:219, note
15) . The specuar, refective distance is postulated as already
foremarked in the fctional state of nature where Rousseau is free to
set up his scene as he pleases; as in all other instances, the diferen
tial relationships that will become the articulations and tensions o
the historical world are already present "en creux" and in the guise of
equalities, in the so-called state of nature. Primitive man is alone and
has no conception of the other whatever, yet already in this absolute
and inconceivable state of solitude, he can b the spectator, the
concern and the judge of his ow singular being. In the apparent
identit of the nominative with the refexive "I," the diferentiation is
foreshadowed and the grammatical space for the futue diferences
staked out. The statement that corresponds to this condition would
have to be 'je m'aime donc je suis," in which the word "aimer" is
needed to posit a refexive structure that the verb "to be" cannot
provde (it is impossible to say 'je me suis") and thus to produce a
genuine cgt, the constitution of a self defned in its ow self
identity by a refexive act. The description of amour d soi pointedy
avoids the scene that would correspond to the nonrefexive, transitive
sentence 'j'aime donc je suis," which would assert the transparenc
ofthe self to its own experience of selfood, the unmediated presence
of the self to itself If the notion of transparenc is at all operative in
Rousseau, it is certainly not associated with the concept of selfood
defned as "amour de soi."
Selflove does not enter the dramatic world of Narcise, which
owes whatever comic efects it can muster to experiences at the
furthest remove from "amou de soi." To a large extent, Narcis e
exploits the hackeed comical resources of vanit, of the amour
propr that Rousseau sharply distinguishes from self-love. In contrast
to the solitar self-concentration of self-love, amour prpre is entirely
directed towards the approval of others. It is a false consciousness, a
mauvaiefoi based on the aberrant assumption that the self, as such,
is worthy of bing loved by the world at large: "on m'aime" or 'je suis
aimable" would b the correct paraphrase for amour propre as op
posed to the ')e m'aime" of self-love.
A false consciousness of this knd can perfectly remain within
166
ROUSSEAU
the confnes of an organizing self that understands and controls its
dyamics. The subject may be blinded by vanity and prone to sy
outrageously false things about the world, yet it remains a subject
and can be kow as such. "What pleasure it will give me to mae
Angelique happy! " exclaims ValEre on the eve of his wedding day;
confronted with his ow portrait disguised to look like a woman, he
reacts by sying "I fnd much in her of my own countenance . . . On
my word, she is charming . . . Ah! if only there is a mind to support
all this . . . But her taste bars witness to her intelligence. The girl is
an expert in the merits of men!" ( 1 :98) . This is certainly a grievous
misreading of the text of the portrait in which some of the structures,
the concern with a correspondence beteen outside attributes and
inner qualities (e
p
rt) , the resemblance acting as a support for the
mystifcation, sugest a pattern ver close to metaphorical substitu
tion. Yet , on the level of amour
p
r
p
r, the portrait is neither a
metaphor nor any other trope.
The mere existence of an aberrant substitution is not sufcient
to set up the specifc complexity of fgation , as long as the error can
be reduced to the intelligbility of an intentional act. On the level of
identifable bad faith the fantastic image that originates in the mind
of the subject and that blots out the world can simply be identifed
with the mode of consciousness that created it ; the reductive reading
of the situation is also the correct one. Once the assumption is made
that the character is vain, the knd ofaberrations to which it is to fall
pre are entirely predictable and the author's skill will consist only in
the invention of more or less surprising situations in which the pre
dictable reaction will choose to ft. The semantic pattern is straight
forward: the misreading of the portrait as being the image of a
pretty girl simply means: Valere is vain. The function of the portrait
is to reveal a consciousness, and since this consciousness turns out to
be a false one, it also seres the corrective function of revealing this
falseness to the subject in the hope that he may mend his ways. The
purpose of the portrait is satirical and didactic and the comical
efects are of this general type. They assume an audience in willing
complicity with what is being done to someone else, for his ow
good. This "good" is not in question, since it is clear that vanity is a
ver stupid way of being-stupid enough, in fact , to degade a man
into a woman ("Yes Valere, up till now it [the portrait] was a woman;
but 1 hope that henceforth it will be a man . . . " [ 1: 1015] ) . Hence, at
all times, the possibility of exploiting ridicule for low comic efect :
"Valre: Wat pleasure it will give me to make Agelique happy!
SELF ( PYGMALION) 167
Fontn (his serant) : Do you intend to make her a widow?" ( 1 : 982) ;
"Valre: Well now, would Mr. Frontin recognize the original of this
painting Fntn: Phoo! You bet I would! A couple of hundred kicks
in the ass and as many blows I have had the honor of receivng fom
him have made it a ver solid acquaintance" ( 1 :985) . In these ex
changes bteen master and serant we are far removed from Sa
gnarelle, or Diderot's Jacques, or Keist's Sosias in Amphitn. The
efects don't have to be that crdely obvous: within some of the fner
moral distinctions made in the satirical scenes, commentators have
found Russeau's play worthy of havng been edited by Marivaux. But
whether crude or subtle, the referential status of all these passages is
unproblematic. The are statements about a consciousness consid
ered as a single unit of meaning and susceptible therefore of being
either "tre" or "false"; no epistemologcal tension or intra-textual
play is involved. On this level , the portrait resembles the deliberate
word "homme" rather than the spontaneous "giant" in the parable
from the Esay on th Orgn ofLngg. But whereas "homme" is
the manipulation of a fgure, the portrait is simply the representation
of a consciousness. The representation of a misreading is not itself a
misreading unless the status of the representation is being ques
tioned as such. As long as we remain in the sphere of amour prpre
this is not the case: vces of this tpe can be "show" without havng
to be "told." One can read Rousseau's entire Confsion as ifit were
such a mimetic histor and, as Starobinski and other commentators
demonstrate,6 it is possible to be both subtle and astute in the pro
cess.
The question arises whether the theme of amour propre (and in
this case it would be entirely legitimate to spea of a "theme") fully
accounts for the dramatic and linguistic efects of the text . Ifit does,
then there is no reason to read Narcise as if it were the dramatiza
tion of a linguistic fgure. It is simply the mimesis of a fawed self
with the existence of the faw allowng for whatever dramatic ten
sion the play contains: the juxtaposition of a mystifed character to
the dear-sighted author and audience, the gradual or sudden discov
er of the subterfuge (anagnorisis) , etc. As such, the portrait is not a
metaphor but one element among others in a human histor of
general psychological and social interest.
There remains however a residue of complication that cannot
6. Se espcially "L progres de l'interpr
e
te" in Jean Starobinsk, L reltin
crtiu ( Paris, 1970) , pp. 8-173.
16
ROUSSEAU
be accounted for i n these terms. Narise contains a number of
grammatical plays on the rfexive mode that point to the greater
degree of generalit also sugested by the Odian title. It contains
sentences of some pronominal intricacy such as "il se cherche pour
s'epouser" (1:1005) , or 'je ne veux plus l'aimer que parce qu'il vous
adore" (1:1016) , or "il vaut encore mieux n'aimer rien que d'etre
amoureu de soi-meme" ( 1: 1002) , as well as plays on the status of
identit that seem to exist for their ow sake, regardless of the vanit
that is supposed to inspire them. The text also sugests that the
introduction wthin the exsting situation of the ambivalent
portrait-part male, part female; part image, part object-reates a
disturbance that no one is entirely able to control , the slightly un
canny "vent de folie" (Baudelaire) that hangs over some of the
scenes, especially those most prominent in the use of linguistic and
hermeneutic terminolog: considerations on "galimatias," "com
prendre" and "expliquer" contrasted in an opposition that is not
obvious ("Luin: Comment m'expliquer ce que je ne comprends
pas" [ 1 : 1006]) , and fnally, of course, the "portrait metaphorise"
which is supposed to explain the incomprehensible (1: 1006) .
The actual narcissistic moment when Valere falls in love with
his ow image is not a moment of pure amour propre. It comes
closer to the situation summarized by Frontin: "il est devenu
amoureux de s ressemblance" ("he fell in love wth his re
semblance" [ 1: 1006]) , a moment suspended between self-love and
the transitive love of others, not quite 'je m'aime" or 'J'aime X" but
rather 'je m'aime come si j'etais X." The self/other tension, latent
when the feelng, as in the Scond Dicourse, is that of pit, has
become objectifed in an autonomous entit, the portrait , that is not
entirely fctional but exists in the mode of a simulacrum. The portrait
has been substituted for the refexive pronoun in 'je m'aime" and it
can do so because it is and is not the self at the same time. It both
resembles the self sufciently to allow for the possibility of self-love,
but it also difers enough from it to allow for the otherness, for the
"pieuse distance" (Valer) that is a constitutive part of all passion.
Valere (not Valer) could just as well be in love wth diference as
with resemblance; resemblance is "loved" because it can be inter
preted as identity as well as diference and is therefore unseizable,
forever in fight.
In this situation, while Valere, like the legendar Narcissus, is
suspended in frozen fascination before an image, the structure is a
SELF ( PYGMALION) 169
geat deal more complex than in the amour prpre of mere vanit.
The portrait is a substitution, but it is impossible to say whether it
substitutes for the self or for the other; it constantly vacillates be
tween both, exactly as in the condition of fear, one constantly vacil
lates between. the suspicion that the reassuring outside might or
might not conceal a dangerous inside or, in the opposite situation,
that the frightening surface may or may not be appearance rather
than evdence. Wereas, in the case of fear, the substitutive oscilla
tion occurs within an inside/outside polarit, in the case of love the
polarities involve a subjective as well as a spatial model and the
fluctuation occurs between self and other, between ipseit and alter
it here reduced to the empirical polarit of man and woman. Wen
it is revealed that the image is indeed the portrait of a spcifc prson,
the ambivalence disappears and mere tricker remains. The interest
ing moment, however, when the play says something signifcant
about the nature and the strcture of love, is when the protagonist
remains suspended between his desire for the portrayed other and
his seduction by reflected likeness.
Love, like perfectibilit, is structured like a fge of speech. The
portrait allows for the bizare substitution of self for other, and of
other for self, called love. From the moment Valere/Narcisse gives in
to this fascination, he considers the portrait "beloved" and trans
forms the suspended vacillation into the defnite identit of an other.
The pattern rns parallel to that of the fctional primitive man in the
Esay who, upon encountering another man, ended the uncertaint
of his feeling by attributing to him the size corresponding to his
suspicions. To the extent that the portrait partakes of amour d si,
albeit in the displaced version of an imagined other, it is indeed a
fgre (the metaphor of a metonymy) , a substitution based on a
refected (contigous) resemblance and leading to aberrant referen
tial conclusions. And since the ver notion of selfood is gounded in
self-love, the loss of epistemological stability in the fgure corre
sponds to a loss of authorit in the self, now reduced to ontological
nothingness: "11 vaut encore mieux n'aimer rn que d'etre amoureux
de soi-meme" (my italics).
This sentence sums up the denouement of the action. Afer
Valere has had it spelled out to him that he was the victim of a
mystifcation, we seem to end up in the transcendence of narcissistic
self-fxation by a normal feeling of love for another. One can expect
the reflexive to become transitive and '1e m'aime" to become simply
170
ROUSSEAU
'J'aime Angelique. " But Rousseau dos not allow for the tranquility
of a transitive world. After havng been "cured" by a public humilia
tion that reduces him to near-oblivon, Valtre reafrms in fact his
self-love, though in a somewhat changed mode: ''nge1iu: Was I
wong in telling you that I love the orgnal model for this portrait?
Val: As for me, I wll consent to love him only because he adores
you" ( 1 : 1017) . From "on m'aime" to ')e m'aime comme autrui ," we
have come to ')e m'aime aimant," apparently one step futher away
from amour prpre toward amur d si. The fnal metamorphosis
of the portrait into Agelique completes the movement that began
when Valere was transformed into a woman. The disrptive portrait
has been domesticated into a reassuring, extra-textual presence, just
as the threatening giant was tamed into "man. " 'je m'aime aimant"
preserves a reflexve moment and combines a referential with an
intra-textual strcture that cannot be called self-referential , since the
"self' to which it claims to point is in fact itself an infnitely deferred
condition of indeterminac beteen self and other, beteen identity
and diference. To the extent that she is inscribd wthin this struc
ture, Agelique's apparent master over the situation becomes illu
sor, as will be evdent when the same situation, considerably de
vloped and enriched, will be treated from her perspective in}ulie.
The return to order, normalc, and proper identity at the end of
Nars is therefore treated as a boufonerie, especially with regard
to the authority fgure of the father, Lisimon. B getting "beond"
amur prpre, a much more disturbing structue is revaled, since it
now bcomes forever uncertain whether the beloved is in fact a
person or a portrai t , a referential meaning or a fgure. "It is better
still to love nothing than to be in love wth oneself' becomes a ver
odd maxm, since "oneself' can never know whether it is anything at
all , and since bing in love with this non-being is precisely the aber
rant hyostasis of a "nothing" into something, delusively called a self.
The transitive displacement of selfood upon the other contaminates
the other's referential identity and opens up the possibility that she
(or he) too is a "nothing." Once we realize that selfood is not a
substance but a fgure, the either/or choice set up in the sentence
loses all meaning, since it then becomes impossible to distinguish
between self-love, the love of ot hers, or the love of nothing. The
rhet oric of self, like the rhetoric of "man" shows that the politics of
love, like the politics of histor, are rooted in the quickands which,
as}uli will show ( or tell) , make up the ethical judgments governing
the relationships between self and other.
SELF (PYGMALION) 171
One other element in the problema tics ofselfood was omitted from
the consideration of Narise and, indeed, from the play itself. The
portrait did not fall from the sky, but had to be painted by someone
skllful enough to hide and to reveal , at the same time, the identity of
the model. 7 From the action of the play we can surmise that the
portrait was commissioned, or painted, by Valere's sister, but this
even t does not enter the text in any way. If however we consider
Russeau as the author of a text in which an action is being "por
trayed," then we can call him the "painter" of the scene in the
classical sense of ut pictor poesi. The later Preface to Narise
( 1753) has much more to do with the reception of the Dicourse
than with the play, but it contains several statements on the self
conscious activty of writing and invites the transference of selfood
from the fctional character into the author who invents and repre
sents i t. By the same token, the indeterminac ofthe self may well b
resolved since common sense tells us that the representation of an
error is a necessar step in its correction. That Rousseau is not or no
longer Vahre is clear from the fact that he is able to paint Valere's
portrait as well as the scene of the portrait's efect upon others. Does
it follow that, unlike his Narcissus fgure (which may be a former
incarnation of his deluded self now lef behind) Rousseau, as a
writer, can be called a selfin a more inclusive sense than the particu
lar and empirical self lost in the confusion of his everday existence?
The Preface to Narcisse seems at least to allow for the possibil
ity: "I confess that there exist a few sublime minds [genies sublimes]
able to dispel the veils with which truth hides itself, a few privileged
souls who are able to wthstand the stupidit of conceit , the low
jealous and the other passions generated by literar ambitions . . .
If any doubt remains as to the justifcation [of my literar vocation] , I
boldly proclaim that it is not with regard to the public or with regard
to my opponents; it is only towards myself, for only by obsering
myself can I judge whether or not I can include myself among the
small number . . . I needed a test to gain complete kowledge of
myself and I have taken it without hesitation" (2: 970-73) . Writing
seems to be held up and justifed as a way to recuperate a self
dispersed in the world. Such a recuperation can b conceived quite
pragmatically or even therapeutically, but it can also be formulated
in more inclusive terms: "What distinguishes Jean-acques from an
7. "Mais Ie portrait exige un artbte" (Starobin:k, jean-aqus Rusau: L
trnsparence et l'obstacl, p. 215) .
172
ROUSSEAU
ordinar neurotic," writes Starobinski, "is that the phantasm, far
from staying self -enclosed, demands its own development in atul
labor [un travail rel ] , provokes the desire to wite, wants to seduce
the public, etc. The decision in favor of the experience of immediac
[Ie parti pris de l'immediat] becomes a literar work and betrays
itself in the process of its manifestation . . . Russeau is projected, in
spite of himself, in the practical, mediated world and one has to
grant that, at least in the case of this extraordinar man, the patho
logical regression of instincts and desires is not incompatible with
the prgsin of his thought."1 This vew is not without its prob
lems, for it is not a prr clear how speculative witing can be called
a "travail rel " ; the oddness of its status as a commodity is apparent
from the erratic relationship between efort and value that governs
the economy of its prouction and consumption. More interesting
still is the suggestion that the experience of immediac can be trans
formed into a literar work that replaces the fulfllment of an ideal
vision by its representation. For this is precisely what Russeau con
stantly warns us against as the most dangerous of aberrations: the
danger of a lapse into a literal version of the state of nature is the
main assertion of the Scon Dicourse. Starobinski himself is anyway
sayng something more complex since (at least in this passage) he
dos not describe the unmediated vision of pure transparenc as a
reality, nor even as a desire, but as a pari pr, the willful assertion
of a likely aberration as well as a resignation to the possibility of this
error. The gesture introduces an important nuance in the strategies
of ego recuperation. It echoes the awareness, in Narcis e, that the
referential representation of what Rousseau calls a passion (as op
posed not only to a need but also to a vce such as conceit) is in fact
the representation of a rhetorical structure which, as such, escapes
the control of the self. Beyond this point, which is when "wrting"
can be said to begin its labor, one no longer starts out from a passion
but fro
m the assertion, the pari pr, of its nonsubjective, linguistic
structure. But is this not the best way to reintroduce the authority of
a self at the far end of its most radical negation, in the highly
abstracted and generalized form of a deconstructive process of self
denial? The statement of the enigma that gives language its necessar
ily referential complexity might itself be no longer a representation
but a single voice that, by the rigor of its negativit, fnally coincides
with what it asserts.
8. Starbinsk, jean-acqu Rusau: L tranparne et i'obtl, p. 215.
SELF (PYGMALION) 173
The rhetorical resouces of language, regdless of whether one
considers them as mere tropes or extends them to wider patterns of
persuasion, are by no means, in themselves, incompatible with self
hood. On the simplest pragmatic level, the obviously ofer the self
the means by which it can accomplish its own designs, either in full
kowledge of its purpose or with the true intent hidden from the
subject by bad faith, repression, sublimation, or whatever dyamics
of consciousness one wshes to imagine. Rhetoric all too easily ap
pears as the tool of the self, hence its perading association, in the
everday use of the term, with persuasion, eloquence, the manipula
tion of the self and of others. Hence also the naIvely pejorative sense
in which the term is commonly used, in opposition to a literal use of
language that would not allow the subject to conceal its desires. The
attitude is by no means confned to the popular use of "rhetoric" but
is in fact a recurrent philosophical tapas, a philosopheme that may
well be constitutive of philosophical langage itself. 9 In all these
instances, rhetoric functions as a ke to the discover of the self, and
it functions with such ease that one may well begn to wonder
whether the lock indeed shapes the key or whether it is not the other
way round, that a lock (and a secret room or box behind it) had to be
invented in order to give a function to the ke. For what could be
more distressing than a bunch of highly refned kes just lyng
around without any corresponding lock worthy of being opened?
Perhaps there are none, and perhaps the most refned ke of all , the
key of keys , is the one that gives access to the Pandora's box in which
this darkest secret is kept hidden. This would imply the existence of
at least one lock worthy of being raped, the Self as the relentless
undoer of selfood.
Within the epistemological labyrinth of fgural strctures, the
recuperation of selfood would be accomplished by the rigor wth
which the discourse deconstructs the ver notion of the self. The
originator of this discourse is then no longer the dupe of his own
wishes; he is as far beyond pleasure and pain as he is beyond goo
and evil or, for that matter, beyond strength and weakness. His
consciousness is neither happy nor unhappy, nor dos he possess any
power. He remains however a center of authority to the extent that
9. Jacques Derrida's "La myhologie blanche" (in Mare d / philoophi
[ Paris, 197], 247-324) is the most powerful recent restatement of this assertion.
Among the numerous antecedents that are bth the target and the confrmation of
Derrida's essy, Knt's rmark on rhetoric in a footnote to the Third Crtiu (Ktk
tr UTiLkrj, 217, note) is a typical example.
174 ROUSSEAU
the ver destrctiveness of hi s ascetic reading testifes t o the validit
of his interpretation. The dialectical reversal that transfers the au
thorit from experience into interpretation and transforms, b a
hermeneutic process, the total insignifcance, the nothingness of the
self into a new center of meaning, is a ver familiar gestue in
contemporar thought , the ground of what is abusively called mod
ernity. Thus in his book on interretation Paul Rcoeur casts Freud
into the role of the rhetorical undoer and the hereneutic recoverer
of the self. Freud has divested the self of any intuitive stabilit what
ever: "all that Freud says [of the self assumes the forgetting and the
vacillation of the self. I O Instead, the self is a mere factor operating
among others in the sstem of quantitative forces and ratios that
Freud's dynamic metaphors describe: " . . . never does the con
sciousness of the Ego appar in Freud's system in the capacit of an
apodictic position, but only as an economic function. "1 1 However, at
the moment that all seems lost , all is regained: "This realism [which
replaces consciousness by an economy] is unintelligible if considered
by itself; the abdication [dsaiismt] of consciousness would be
properly speaking sels [inen, meaningless but also insane] , if
its only result were to alienate reflection into the consideration of a
mere thing. This is what would happen if we omitted the complex
ties that bind this topico-economic explanation to the efective labor
of interpretation, which makes psychoanalysis into the decphering
[dcrptage] ofa hidden meaning into an apparent meaning. "12 The
Freudian deconstrction is only a necessar prelude to a "recover of
meaning in interpretation" 1 3 and the subject is reborn in the guise of
the interreter: "Reali t of the it [the nonsubjective id of the uncon
scious] because the it gives food for thought to the exegete" ("Ralite
du <a en tant que Ie <a donne a penser a l'exegete") . 14
The part here played by Frud (and we are not now concerned
with the ''validit' of this interpretation with regard to Freud) could
equally be assigned to literar texts, since literature can be shown to
accomplish in its terms a deconstrction that parallels the psy
chological deconstruction of selfood in Freud. The intensit of the
interplay between literar and pschoanalyical criticism is easy
10. Paul Rcoeur, D l'intrpretton, E ai sur Fu ( Par, 195) , p. 416.
1 1 . Ibid. , p. 416.
12. Ibid. , p. 416.
13. Ibid. , p. 41 1 .
14. Ibid. , p. 425.
SELF (PYGMALION) 175
enough to understand in these terms. The same strateg occurs in
certain philosophical texts or readings, for example in Heideger,
who also locates the deconstruction of the self as substance in a
hermeneutic activt which, in its turn, becomes the ground of a
recover of selfood as the springboard of futurit: "Wen fully con
ceived, the care-strcture includes the phenomenon ofSelfhood. This
phenomenon is clarifed by interpreting the meaning of care; and it
is as care that Dasein's totalit of being has been defned. " 1 5 Our
present concern is merely whether Russeau, like Ricoeur's Freud,
reclaims a measure of authority for the self, grounded in its abilit to
understand its ow failure to make such a claim.
The consistenc wth which Rousseau dramatizes this ver ques
tion in another Ovidian text staged, like Narcis e, for theatrcal repre
sentation, illustrates the recurrent symbiosis of the problems of
understanding Wth those of selfood, a pattern that is obvious in the
sparse philosophical references we have mentioned but that could be
extended at almost infnite length. The "scene lyrique" Plion
( 1762) , in contrast to the Preface to Narcis e wrtten with the antici
pation of an extensive productive labor ahead, looks back upon a
considerable body of literar achievement, including the invention,
in the fgres of Emile and Julie, of convncing fctional subjects. The
situation of the scene, that of an author confronting his own fnished
work, corresponds to the actual predicament of Rousseau at that
time, just as the position of Plion wthin the Rousseau corpus
marks the transition from theoretical and fctional to autobiographi-
15. &in un Zit (Tubingen, 1927) , p. 323, quoted from Being an Tm,
trarslated by John Macquarrie and Edward Rbinson ( New York, 1962) , p. 370.
Heideger is however a geat deal subtler ( or more devous) than Rcour in insisting,
in reference to Kant, that "the suqjectum is 'consciousness in itself,' not a representa
tion but rather the 'form' of representation. That is to sy, the 'I think' is not
something represente but the formal strcture of representing as such, and this
formal structure alone makes it possible for anyhing to have been represented."
(Sin un Zit, 319; Being an Tm, p. 367) . The "formal strctue of representa
tion" ( prhaps wth the omission of "as such") is what we call rhetoric or, btter,
rhetorcity. Rcoeur, on the other hand, considers Freud's distinction between a drive
[Tb 1 and the representation of this drive as an unquestioned extension of a valid
realism: "C'est parce que ce realisme est un realisme des "presentations" de pulsion,
et non la pulsion elle-meme, qu'i! est aussi un realisme du connaissable, et non de
l'inconnaissable . . . " ( Rcoeur, p. 422) . The epistemological integity of the rhetori
cal moment remains "inpnse" whereas, in Heideg eer, far from being simply re
pressed, it accomplishes the much more redoubtable feat of bcoming the "totality
of Being."
17
ROUSSEAU
cal works. The fact that the text, as we understand it, aserts in fact
the impossibilit of makng these facile generic distinctions should
caution one against following all too confdently the hints provided
by the convenient evidence of chronolog.
The lyrical scene, as Rousseau calls it, starts out as a confronta
tion between the artist and his creation and culminates in the highly
dramatic moment when the statue of Galathea comes to life. This
moment is preceded and also followed by a great deal of contradic
tor verbal agitation, by no means clear in its dramatic and semantic
function with regard to the central event. No wonder that the brief
text, despite its obvious dramatic shortcoming, has engendered a
minor but distinguished tradition of misreading which includes
Hamann, Shiller, and Gethe. 1 6
The general movement of the text is one of constant vacillation,
explicitly identifed as such, since Pgmalion repeatedly rejects, in a
sequence of dramatic reversals, the understanding he seems to have
acquired of his situation. As was already the case in Narcis e, the
possibility of interpretational error is thematized throughout the
text. Consequently, none of the statements in Pgmalion's mono
logues can be taken at face value; they all function within a
contextual movement that stands itself in need of interpretation but
wthout which their validity in terms of truth and falsehood cannot
even be considered. The provisional syntheses that are achieved along
the way in the course of the action do not necessarily mark a progres
sion and it is the burden of the rading to decide whether the text is
the teleolog of a selfood that culminates in the climactic exclama
tion "Moi ! " or a repetitive vacillation.
The encounter between author and work with which the play
begins-as the parable of language began by the encounter between
two men-ccurs in a complex afective mood. In the scene from the
Es ay, the emotion was fear; in Naris it was love, complicated by
its various self-refective creases; in Palion, it is a combination of
both. The presence of love accounts for the choice of this particular
myth, but the fnished statue, considered in retrospect, also inspires
fear: "I don't kow what I feel in touching this veil ; I am seized by
te
ror (ayeur] . . ." 1 7
16. Se Hermann Shluter, D Palion-Symbol bei Rus au, Hamnn,
&hiller ( Zurich, 196) and his biblioga
p
hy.
17. Palion, texte etabli et annole
p
ar Jacques Scherer in J. J. Russeau,
SELF (PYGMALION) 17
Fear in Plion cannot simply be equated with the discrep
anc between appearance and realit, outside and inside, that or
ganized the scene from the Esa
y
. The threat that occurs here is less
obvious than that of physical assault . It is rather the paralysis that
aficts Pgmalion and that reduces him, at the onset of the action, to
a state of dejection. Consequently, the metaphor that it generates wll
have to be further-reaching than the "giant" invented b primitive
man. Pgmalion is paralyzed by the feeling of awe that is character
istic, to use Kntian terminolog, of the sublime. The threatening
power is not something exterior that one confronts directly in an
unmediated encounter: it has instead been transferred, by an act of
the mind (sometimes called imagination) into the constitution of an
entit, a subject, capable of refecting upon the threatening power
because it partakes of that power without however coinciding wth
i t. IS Awe is not directed towards a natual object since it actively
involves the self, 1 9 nor is it directed towards something that could
conceivably coincide with the self (such as "man") , since such an
equation would be merely evasive. Russeau takes his metaphor
from the myh in which the work of art is presented in the gise of a
godess: " . . . I feel as if ! were touching the sanctuar of a goddess
. . ." ( 1 : 1226) . The awesome element in the work of art is that
something so familiar and intimate could also be free to be so radi
cally diferent . Unlie nature, where the diference is easily concep
tual ized into a dichotomy of subject and object , the work of art exists
as a nondialectical confgation of sameness and otherness, suf
fcien tly uncanny to be called godlike.
The goddess metaphor is an aptly monstrous concatenation of
self and other. Galathea partaks of divinit not because of her objec
tive beaut, the Phagorean harmony of her proportions; the rich
iconogaphy of the topic is uniformly ludicrous or at best , as in the
case of Falconnet , banal. In the Rousseau text , her beauty is notice
able only in the emotional gesticulation of her maker. Her godlike
qualit stems from the discrepanc between her specular nature, as
an act of the self in which the self is bound to be reflected, and her
Ouv complt, ed. Berard Gagnebin and Marcel Ryond ( Pars: Gallimard
[Bibliotheque de la Pleiade], 1961 ) , 1 : 122-31 .
1 8. Immanuel Knt, Kri tik t UiLkraf (Stuttgrt, 1 9) , 2, pp. 105 i
19. Knt, 2 ( 105) : "Thus, in our aesthetic judgent, natue is not consid
ered to b sublime bcaus it is awesome but bause it awakens our pwer (which
is not that of nature) in ourslves . . . "
17
ROUSSEAU
formal natre which has to be free t o difer from the self as radically
as can be imagined. This discrepanc prouces the system of an
tinomies that confront each other at the beginnig of the play,
evoked by such tradi tional plarities as hot and cold ("all my fre has
been extinguished, my imagnation is ice cold . . . ") , art and nature
(" . . . masterpieces of natue that my art dared to imitate . . . ") ,
man and god ("Pgmalion, make no more Gds: you are only a
common artist . . . ") . Those antinomies are not rooted in natural
oppositions but are coordinated in terms of the relationship between
self and other that engenders them. This is true even of the most
natual and "material" polarity to appear in the text , that of hot and
col d.
The statue i s "cold," sheer marble not because it i s made of
stone, but bcause, from the ver beginning, it reflects the fgural
coldness of Pgmalion's condition: "All my fe has been extin
guished, my imagination is ice cold, the marble comes cold out of
my hands . . . " 1 : 1224) . The original , literal coldness of the marble
had been turned into fgural heat at the moment of invention, and
this heat had in its turn fed the enthusiasm of Pgmalion as he
engaged nature in the analogical process of imitation, in which the
common properties of art and of natue, of "genie" and of "amour,"
are revealed and exchanged; nature can then b addressed as " . . .
masterpieces that my art dared imitate, . . . beautiful models that
fred me with the ardr of love and of genius . . . " ( 1 : 1225, my
italics) . But when the action starts, we are well beyond the illusion of
this "beautiful" balance between self and other. The exchange breeds
its own excess, engenders its own subl imit, as the illusor presence
ofthe self gives the object quasi-divine powers which, in turn, reduce
the subject to the awestruck baflement of a wll entirely alienated
from its work. This knd of awe, frozen bfore its own sublimit, is
again called cold (by inferential contrast to the "ardor" of the im
itated models) : " . . . masterpieces of nature that fred me with the
ardor oflove and genius, since I surpassed you, you leave me entirely
indiferent. " But this coldness has nothing in common with the cold
ness of the original stone; Bachelard' s thermodyamics of the mate
rial imagination would fnd nothing to feed on in Plion. "Hot"
and "cold" are not , in this text , derived from material properties but
from a transference from the fgural to the literal that stems from
the ambivalent relationship between the work as an extension of the
self and as a quasi-divine otherness. This was implicity asserted as
SELF ( PYGMALION) 17
soon as Galathea is sid to b godlike; from the moment the sublime
is involved, nature recedes in the background and, as in Knt , a
terminolog of selfood an self-consciousness taes over. 20
The divinity also has to be a god, an object of erotic desire.
Ambivalences of self and other are thematized in terms of "love" and
Pgmalion's awe contains self-erotic as well as transcendental ele
ments. Galathea turs into Venus and Pgmalion's erotic self
fascination is similar to Valere's in Naris: je m'aim aimant is
echoed in Pgmalion's 'je m'adore dans ce que j'ai fait . . . " ("I
worship myself in what I have produced" [ 1 : 1226]) . Except that the
object of his love is not just any woman, not even an angelical one,
but a goddess: the sublime dimension is the product of a self awed
by the kowledge that he is the agent of his ow production as radi
cally other. If, in the passage from which this sentence is quoted, this
selfidolatr ( 'e m'adore" instead of 'e m'aime") is called amur
propre ('je m'enivre d'amour propre" [ 1 : 1226]) , it is only related by
h
y
erbole to the mere vanity for which men of letters are so fre
quently being blamed in Russeau's writing. <. 21 Pgmalion's fascina
tion is of a diferent order and the social satisfactions of recogition
no longer can touch him: ". . . praise and glor no longer elevate my
soul ; the appreciation of those who will be cherished by psterity no
longer touches me; friendship even has lost its appeal for me"
( 1 : 122-25) . For the self-idolatr, considered from the perspective of
the subject , is by no means a mere mystifcation: the inventive power
of the self is trly uncanny in its escape from any willful control .
There is no limit to the wealth of its discoveries and one must
assume that , for the author of Juli, the surpriss of self-reading are
inexhaustible.
The ambivalence of this scene of reading ( or wrting) carries
over into the structure of its representation. The scene is both static,
with Pgmalion locked into the fascinated concentration on a singe,
perfect object at the exclusion of anything else, and animated by the
restlessness of a desire that disrupts all tranquil contemplation.
20. Knt, 2 (10): "Thus the sublime is not contrived in nat ural thin@ but
in our consciousness [ Gemut ] . . ."
21. "L gout des lettre . . . nait du desir de se distinguer . . . Tout homme
qui s'ocup des talents ageables veut plain', etre admire, et il veut etre admir plu
qu'un autr. Les applaudissments publics appartiennent a lui seul:je dirai qu'il fait
tout pour les obtenir, s'il ne faisait encore pl us pur en priver ss concurents .
"
( Preface de Naris e, 2:95, 967-6) .
t80
ROUSSEAU
Pgmalion's fetishism of selfood, inhernt i n his vocation as "ar
tist ," is an unstable juxtaposition of plenitude and disruption. The
predicament is represented in the absurd gesticulation of a man
caught wthin the space that immediately surrounds the statue yet
unable to remain within it : "Pgmalion . . . dreams in the attitude
of a restless and sad man; then suddenly rises . . . "; "Pgmalion
walk dreamily around for a while, . . . sits dow and looks arund
him . . . then rises impetuously" ( 1 : 122-25) . The same tension is
evoked less naIvely in the paradox of an excess that is also a lack, the
supplementar structure that Derrida has so accurately described in
Rousseau. The perfection (". . . never did anything so beautiful ap
pear in nature; I have surpassed the handiwork of the Gds . . .")
stems from an excess that sets the statue

apart from the merely
natural world; the excess in turn engenders a lack ("Ah! her only
shortcoming is her perfection"-"c'est la perfection qui fait son de
faut" [ 1 : 1227]) , and there is no escape from the pressure ofa difer
entiation that never allows for a totalizing integrit. The scene has to
get in motion, and the initial polarities that had been frozen in static
opposition have to enter into a play of substitutions and reversals.
The lyricism of the opening monologue has to become dramatic,
turn into the mied genre ofa "scene lyique," the representation ofa
godlike self, Dionysus on the stage.
The text is dramatically strctured as a dyamic system of ex
cess and lack (defaut) metaphorically represented in a polarit of self
and other that engenders, in its turn, a chain of (as)syetrical
polarities: hot/cold, inside/outside, art/nature, life/death,
male/female, heart/senses, hiding/revealing ("Et toi , sublime es
sence qui se cache aux sens, et se fait sentir aux coeurs . . ."
[ 1 : 1228]) , eye/ear (in the apparent progression of the text from seeing
the statue to haring it speak in its ow voice) , lyric/dramatic, etc. In
a passage like the following, the antinomies achieve intense conden
sation: "All your fres are concentrated in my heart and the cold of
death remains on the marble: I perish by the excess of life that [the
statue] lacks . . . Yes, two beings are lacking from the plenitude of
things . . . " ( 1 : 1228) . Systems of this tpe would evolve harmoni
ously by means of exchanges of properties if they stand under the
aegis of a totalizing principle ( here called ' sublime essence," "soul
of the universe," "principle of all existence," "sacred fre," etc. ) that
fun
ctions according to a balanced economy, the rich spontaneously
giving to the poor because benevolent nature's law is that of a dis-
SELF (PYGMALION) 181
tributive justice: "the natural order has been upset , nature is out
raged; give back their strength to nature's laws, restore its benevolent
fow and distribute your divne infuence with equanimity . . . "
( 1 : 1228) . At this point in the action, these statements are 'made by
Pgmalion at the height of his self-mystifcation, in the tone of a
prayr that receives a by-no-means unambiguous answer. What the
passage proves is that Russeau controls the rhetoric of totalization
inherent in all supplementar systems, from its most naive to its
most devious forms. Unavoidably, these are precisely the statements
that will be retained as the commonplaces that become the mains
t
ay
of traditional intellectual histor: equally benevolent natual
economies will soon inspire Rousseau's readers, such as Herder for
example,22 and it will be almost impossible to escape from their
seduction. When this fnally happens, it is equally predictable that
the author blamed for the aberration which he identifed only in
order to denounce it , would precisely be Jean-Jacques Rousseau.
The processes of substitution at work in Plin are not all as
transparently wishful as in the passage that has just been quoted and
that culminates in Pgmalion's pathos-flled apostrophe to Galathea
( 1 : 1228) . The text achieves a higher degree of dialectical complexty.
For exmple, the dialectics of desire are allowed to develop along
consistent lines in the frst part ofthe scene as the evolve from literal
sexual agression to the idea of the "beautiful soul" to the dialectics
of the general and the particular and fnally to the apotheosis of the
self by its immolation to the work.
Goethe would have grievously misread Pglin if the com
plaint that Rousseau "wanted to destroy the highest that mind and
deed can produce by the most common act of sensuous lust
,,
23 were
to be taken seriously. Pgalion's gesture ofliteral sexal aggression
is comically transparent : "He takes his mallet and his chisel and
hesitatingly ascends the steps leading to the statue. He seems afraid
to touch i t. Finally, with his tool already raised, he stops. 'Such
trembling! How troubled I feel ! I hold the chisel with faltering hand
. I cannot . . . I dare not . . . I would spoil everhing' "
22. In his Ahnlung iber dn Urprung dr Sprah, Herder refers to the
"unitng laws of Natue's econom [d Vebinunggeetze dr hauhltrn
Natur]. The complications of Herder's ow text remain to b analyzed, in themselves
as well as in relation to Hamann's subsequent attack.
23. Gothe, Dihtung un Wahrhit, Part 3, Bk 11, in Gth Werk, Ham
burger edition, eite b Lieelotte Blumenthal (Hamburg, 1955) 9:49.
18
ROUSSEAU
( 1 : 1226) . But sexual assault i s by no means incompatible with the
aesthetics of the sublime, and Gethe's polarit of high and low, of
"the highest" opposed to "the most common," entirely misrepresents
the signifcance of the scene. The agession, reminiscent of Saint
Peux's fetishistic fantasies upon receivng Julie's portrait Uuli,
2:278-79) , simply reafrms the referential efectiveness of any
metaphor. Love as agression, Venus and Mars, is a necessar
thematic projection of any pattern of metaphorcal exchange, since
the representation of copulation or of muder are the most efective
emblems for the moment of literal signifcance that is part of any
sstem of tropes. Desire is built into the system, the ony problem
being whether it can be fulflled by an exchange that would put a
stop to the endless chain of substitutions. The text of Plin
sugests a general version of such a closue and this part of the play
is ordered as a dialectical progession. It starts in the inside/outside
analogism of the "beautiful soul" ("How beautiful must b the soul
to animate such a body!" [ 1 : 1227]) in which the intrinsic qualit of
the selfis borrowed from the surace ofits physical shape. The literal
sexual agression becomes the "sybolic" adoration of Galathea's
soul , a displacement rather than an intensifcation of the initial er
ror, since it was never sugested that the statue was a substitute for
an actual person in the fst place; Galathea's divne status is
afrmed from the start and is constitutive for the text. Pgalion's
problem is not that a desire for a particular person is idealized and
called divine but rather that the abstraction and the generalit of a
linguistic fgure manifests itself necessarly in the most phyical of
modes. In movng from the delusion of un mediated possession to the
delusion of a corespondence between body and soul , we have
neither progessed nor regessed towards a futher degee of truth or
falshood, but merely moved to another place wthin the strctue.
Pgalion discovers the metaphor of the beautiful soul to be as
unsatisfactor as the referentialit of the body; rather than copulat
ing with literal meaning, it is more akn to copulating with a stone:
". . . so it is because of this inanimate object that I am unable to
leave this room! . . . a piece ofrble! a stone! a shapeless and hard
piece of matter, worked on with this iron tool! . . . Fool, come back
to your senses, lament your fate, see you eror . . . see you mad-
ness . . . But no . . . " ( 1 : 1227) . By conceivng of the soul analogi
cally, by way of its assumed resemblance to a materal entit, one
has in fact reduced it to being one more stone among stones. But this
SELF ( PYGMALION) 18
moment of negative insight is at once reintegated within the pro
cess, thus allowing for a "higher" couple of antinomies to come into
play: "No, I have not taken leave frm my senses; no, I am not
delirious; no, I don't have to blame myself for anything. It is not this
death marble that I love, it is a living being that resembles it; I am
entranced by the shape (gre] that it presents to my eyes. In what
ever place this adorable shape may be, whatever body may carr it ,
whatever hand may fashion it , I have gven it all my heart" ( 1 : 1227) .
The synthesis now returns to the level of appearances ("la fgre que
l'etre vvant ofre a ms yu
"
) , but it taes place beteen a particu
lar being and the general principle of which this being is a repre
sentation. The inside/outside metaphor of the beautiful soul is re
placed by a sythesis ofthe particular and the general reminiscent of
what neoclassical witers refer to as "general beauty." This aesthetic
"generality" does not correspond to what Rousseau had earlier called
the ''enerl will" in his political wtings and especially in the Scil
Conta;24 here, the term "general" (which does not appear in the
text) desigates a syecdochal metaphor in which the whole has
priority over the part. The genrl model is not a combination of
miscellaneous particular traits (as if Galathea were an amalgama
t ion ofthe various individual women that animated Rousseau's erotic
reveries) , but the attraction of the indivdual stems from its re
semblance to a prior general model that is, in fact , an emanation of
the self. Only indivduals who partake of that general principle can
be beautiful and desirable. Aesthetic generality is the precondition
for resemblance which also means that it is constitutive of metaphor.
One is reminded of Frontin's statement in Narise: "il est devenu
amoureux de sa ressemblance" ( 1 : 1006) . The "progression" from the
literal to the general remains wthin the tropological pattern of sub
stitution that makes Palin into an allegor of fguration. The
various steps in this progression do not simply cancel each other out ;
they are "aufgehoben," surpassed but maintained ("No, I don't have
to blame myself for anything," says Narcisse, in a statement that
nothing in the text disproves) , which does not mean that they are
allowed to reach their teleological closure.
The sthesis of the general and the particular is clear enough
as long as it involves anothr rather than the self. 'Je suis epris d'un
etre vivant qui lui ressemble" is entirely understandable as long as
2. Oro "genera]" in the Sil Conta, see pp. 27-49.
18
ROUSSEAU
the "lui" refers t o another, i n this case feminine, identity. But if, as in
Narcis e, the situation is rather 'Je suis epris d'un etre vvant qui m
ressemble," then the crossing of generalit and particularity with the
categories of self and other engenders the disorder heard in Pgmal
ion's "passionate," "transported" speech. The ambivalence of self
and other is actively at play in the mode ofthe sublime and the claim
to generality has to extend to the self as well . This implies that the
"general" shape of Gala thea's statue i the selfin a radical sense. The
work no longer originates in the particular will that shaped it , but it
is the work that causes the self to exist as its ow source and tlo:
"Wat lines of f seem to emanate from this oqiect in order to
enfame my senses and return with my soul to their souce" ( 1 : 1228) .
The work alone is now the source oflight and life, both miror and
lamp. The work reads the man and reveals his total insignifcance
except in his relation to the work. This apparent immolation of the
self ("Ah! let Pgalion die in order to relive in Galathea! " [ 1: 1228] ;
". . . I have gven you all my being; henceforth I shall live ony through
you" [ 1: 1232]) is in fact its glorifcation, for at its cost and only at
this cost can the work b called a souce and made the center of all
life, the "holy fre" in which only Sages can dance. Poetr draws its
most efective seductions from this temptation, from Blake's myth of
the Book of Fire "displayng the infnite which was hid . . . by
printing in the infernal method, by corrosives" made of "faming fre,
raging around and melting the metals into living fluids" t o be "re
ceived by Men . . . and [to take] the forms of books to b arranged in
libraries. "25
Pgalion, as a character, sweres away from this temptation
as soon as it has been explicitly stated and revealed. In the central
articulation that sets the dramatic pattern for the text , when the
totalizing identifcation is about to occur, the exchange between self
and other that was to abolish all polari ties does not tae place or,
perhaps more accurately, leaves a surplus (or a defcienc) that pre
vents the narration from closing: "Ah! let Pgmalion die in order to
relive in Galathea! . . . Heavens! What am I sayi ng! If I were she, I
would not see her, I would not be the one who loves her. No, let my
Galathea live, and let me not b she. Ah! Let me always b another so
that I may always wsh to be she, to see her, to love her, to b loved
25. William Blae, Th Marre ofHeavn an Hell, in Th Potal Work of
William Blk (London, 1913) , p. 255.
SELF (PYGMALION) 185
by her . . . " ( 1 : 1228) . As a mere reassert ion of "pie use distance" the
passge (except for the high degee of pronominal condensation)
says nothing that was not already stated in Narcisse's fascination
with the portrait . But the location of the statement within the play,
afer the sstematic development leading from the literal to the
analogical to the general has been allowed to unfold, gives it added
signifcance. The totalizing symmetr of the substitutive patter is
thrown out of balance: instead of merging into a higher, general Self,
two selves remain confronted in a paralyzing inequality. The
categories that are thus being challenged are precisely those of self
and other, the ground of the system. If these polarities have only been
posited in order to eliminate their opposition, then the failure of the
synthesis, the persistence of their antagonism, reveals the fallac of
their position. And there can be no doubt about their continued
confrontation, in endless repetition, in the apparent conclusion of the
tet . The fnal exchange between Galathea and Pgmalion reiterates
the situation that existed in the central passage when Pgmalion
wthdraws from ultimate identifcation with the most generalized
form ofselfood. IfGalathea's coming alive (a moment that remains
to be accounted for) confrms this interpretation, then the play
could, in principle, have come to a stop in the identitng echo of the
two "moi's" uttered by the protagonists: "Galtha (touches herself
and says) : Moi. Plion (transported) : Moi ! " ( 1: 1230) . The
supplementar exclamation mark records the imbalance acted out
in the fnal exchanges. Galathea setting herself apart from the mate
rial stone ("Galtha takes a few steps and touches a marble stone: It
is no longer I") is clear to the point of redundanc, but her statement
afer touching Pgmalion is as ambiguous as Alene's famous
"Ach! " at the end of Keist's play Amphiton: "Gala thea goes in his
direction and look at him. He rises precipitously, stretches out his
arms toward her and looks at her ecstatically. She touches him with
one of her hands: he trembles, takes her hand, presses it against his
heart , then covers it with kisses. Galth ( with a sigh) : A! encore
moi" ( 1 : 1230-31) . The tone is hardly one of ecstatic union, rather of
resigned tolerance towards an overassiduous admirer. Since
Galathea is the Self as such, she has to contain all particular selves
including Pgmalion; as a statement of identity in which "encore
moi" means "aussi moi" ("me as well") , it is a true enough afrma
tion. This is certainly how Pgalion understands i t; it provokes in
him the same language of self-immolation that was checked earlier:
18
ROUSSEAU
"Plion : It is you, you alone: I have given you my entire
being; henceforth I shall live only through you. " But the line "Ah!
encore moi" spoken with a sigh that sugests disappointment rather
than satisfaction can also mean "de nouveau moi" ("me again") , a
persisting, repeated distinction between the general Self and the self
as other. Indeed, the separation between Galathea's coldness and
Pgmalion's impetuousness could not be geater, all the more so
since appearances to the contrar nothing has happened that would
diminish the validit of the earlier moment of wthdrawal before
self-destruction and prevent its repetition. It is true that by now the
statue has come alive, but the text is pointedly set up in such a way
that the epiphany does not occur as a reward for the sacrifcial
self-transcendence but only afer the hope for the success of such an
economy of all or nothing has been abandoned: "Plion : . . .
Alas! In the condition I'm in, one invokes all and nothing hears us.
The hope that misleads us is more senseless than desire" ( 1 : 122) .
When Pgmalion returns to a rhetoric of self-annihilation in the fnal
lines, one can assume from the preceding movement of the text that he
has agin been led astray by a hope now know to be unfounded.
Rousseau and his reader, together with Galathea, can now part com
pany frm the character Pgmalion (which was not the case earlier at
the moment when Pgmalion reafrmed his distance) and notice that
the concluding scene is not in fact a conclusion but one more vacilla
tion in a sequence of reversals , none of which have the power to close
of the text .
The separation of the group work-au thor-reader from the con
sciousness ofthe protagonist indicates that we are no longer wthin a
thematic context dominated by selfood but in a fgual representa
tion of a structure of tropes. The coming to life of the statue does not
occur in response to the most advanced stage in the dialectic of the
general and the particular, in the self-sacrifcing negation of the
subject. It takes place afer the "cold" mood which sees throug this
strateg has been allowed to assert itself. That this radical negation
of the selfis in fact its recuperation is evdent from its text-producing
power: even in this brief play, it engenders most of the "heat" that
keeps the language alive and allows for the coinage of the paradoxes
based on binar oppositions. The language of pathos is infnitely
eloquent. However, the strcture of priority represented in the time
sequence of the dramatic action indicates that the moment at which
Galathea fnally consents to bon'Ow from Pgmalion the excess that
SELF (PYGMALION) 187
she lack, she feeds upon colder fres than those burning at sacrifcial
altars. The energ that succeeds at last in forcing the exchange is the
deconstrctive discourse of trth and falsehoo that undoes selfoo
as tragic metaphor and replaces it by the knowledge ofits fgural and
epistemologically unreliable strctue. When Galathea comes alive,
Pgmalion is no longer a tragic fgre but , like Rcoeur's Freud, a
deconstructive interpretative process (a reading) that can no longer
tolerate the pathos of the self. Galathea's coming alive rewards the
access to this advanced level of understanding. The point of the text
however is that even this mode of discouse fails to achieve a con
cluding exchange that would resolve the tension of the original dejec
tion. The part of the action that follows Galathea's epiphany disrpts
the dialectical progression that leads up to it and merely repeats its
abrrant pattern. The discourse by which the fgral strcture of the
self is asserted fails to escape from the categories it claims to decon
strct , and this remains tre, of course, of any discourse which
pretends to re-inscribe in its tun the fge of this aporia. There can
be no escape from the dialectical movement that produces the text .
Narcie revealed the fgal structue of selfood as it operates
in the relationship between subjects. Plion, on the other hand,
represents the more complex relationship between selfood as
metaphor and the representation of this metaphor, the "formal
strcture," in Heideger's words, "of rpresenting" which is also the
main concern of Kant's critique of judgment as aesthetic judgment.
On this level of rhetorical awareness, the previous metaphors such as
"giant" and "portrait" have been extended to bcome the most gen
eral , all-inclusive concept of selfood conceivable: what remains
after any "self' -interested motions of selfood, even at their most
sublime or their most rigorous have been negted. Rousseau's refusal
to gant authority to even this level of discouse, despite the fact that
the dialectical development that leads up to it is controlled at all its
stages, indicates the impossibility of replacing the epistemolog of
fgural language by that of the self. From the point of vew of truth
and falsehood, the self is not a privileged metaphor in Russeau. This
obviously has consequences for the way in which his autobiogaphi
cal texts, from the Lters to Ma
l
eshrbe to the Reries , would have
to be read. 26
26. An attempt in this direction i made in the last chapter on "Excuses. "
9
Aegor
Ouli)
I F THE SE L F I S NOT, I N PRI NCI PLE , A PRI VI L EGED CATE
gor, the sequel t o any theor of metaphor will be a theor of narra
tive centered on the question of referential meaning and not merely
on the pronominal substitutions that organized such texts as Nar
cUse and Pion. Texts will of couse always contain substitutions
of this type but the will not necessarily determine the main narra
tive articulations. Rousseau's most extensive narrative fction, the
novel}uli ou l Nouvell Heloi, contains elaborate substitutive pat
terns involving the polarities of self and other; if this were not the
case, "love" would not be the prominent thematic element that it
obviously still is injuli. But the ver clearly markd division of the
novel in two parts can no longer be interpreted in terms of this
model . What is involved in this division determines one's understand
ing of the text as a whole and requires a further elaboration of the
general theor of fguation established in the Scon Dicourse.
For it is clear that the problematics of fgural language have not
run their full course in such texts as the Scon Dicourse, an
thropological fctions that carr out the deconstruction of conceptual
language. A narrative like juli cannot be reduced to parables of
denomination. If the thematic prevalences are any indication (some
thing which can only be postulated heuristically for the sae of expo
sition) , then the major diference betweenjuli and the Discourse is
the presence, in the former text, of a strongly marked ethical dimen
sion in the foreground of the action. Just as, in the Scon Discourse,
the burden of the interpretation consisted of fnding the passage
from the linguistic structure to the political assertion, the challenge,
in the reading ofuli, is the articulation ofthe fgural mode wth the
ethical tonality.
Much has gone astray in the critical reading ofulie 1 because not
1 . All references to Juli ou l Nouvll Helie are to j. J. Russau, Ouv
com
p1te, edited b Bernard Gagebin and Marcel Ryond (Pari s: Gallimard [Bib-
18
ALLEGORY (JULIE) 189
even the generalized notion of selfoo, let alone the linguistic prob
lem of referentialit, has been recognized. I am not only thinkng of
the literalism that debates the priorities between the assumedly
"real" correspondence between Mme d'Houdetot and Russeau and
the fctional corespondence btween Julie and Saint-Peu, or that
speculates and passes judgment on the psychological verisimilitude
of the mnae a toi at Clarens. Al this is a little like speculating
whether Fichte's absolute Self rides on horseback or stands on a
mountain top, or whether Knt's schemata have blue or brown eyes.
The more Rousseau tried to avoid particularization, for example by
reducing distinctive physical traits to the minimum need for allegor
ical signifcation2 or by makng the epistolar stle almost intolerably
uniform,3 the more readers have felt complled to fll the space thus
allotted to thei fantas with trivia. 4 The fallac of realistic fction
seems to have blinded us to the fgal abstraction invited by the
neo-medieval title, although it should be obvious in a work in which
"characters" have little more human individualit than the theologi
cal virtues, the fve senses, or the parts of the body.
But evn on a much more refned level of critical awareness,
when Juli is read as a novel of inwad self-refection that might
anticipate Alph, say, or Obermn, or even some aspects of
Baudelaire or of Poust , we are still coping with a contingent and
basically irrlevant misreading. For one thing, such a reading keps
consideringuli, ifit considers it at all , as if it would have preferred
it to be the Consin or the Reri rather than what it is. "It is
truly regrettable," writes Bernard Guyon, the latest editor of the
Nouvelle Heli, "that in his dialogue-preface and even in his report
liotheque de la Pleiade] , 1961) , vol . 2, text established and annotated by Brna
Guyon. Guyon's notes bring the earlier critical editions of the Nouvl Heli
( including Daniel Moret's edition of 1925) up to date.
2. Se Rger Kempf, Sur l cr rmnu (Paris, 196) , pp. 49-65.
3. Diderot's characterization of the epistolar stle of Rchardson woud cer
tainly not apply toJuli: "Un homme qui a du gout ne prendra point les lettres de
Mme Norton pur la lettre d'une des tantes de Clrs, l lettre d'une tante pou
celle d'une autre tante ou de Mme Howe . . . Dans ce live immortel, comme dans l
nature au printemps, on ne trouve pint deu feuilles qui soient d'u meme vert.
Qele immens vrete de nuances!" (Eloge de Rchardson, in Dnis Diderot , Ouve
complt, chronological edition with introuction by Rger Lewinter [Pa, 1970] ,
5: 136) .
4. Berard Guyon thinks that the deepst source of Juli is the "demon du
midi" of a man who has pissed the age of fort (Introuction, 2:xi-xii) .
19
ROUSSEAU
on the novel i n the Consion [Russeau] has put such stress on the
diferences that separate the to frst parts from the rest of his novel
. . ." (2:xii) . Something of this "il est vraiment regrettable . . . "
lingers on in much less naIvely reductive readers, who still would
have wished the book to b somehow diferent. The critics most
astutely responsive to the seduction of Russeau's refective inward
ness, Marcel Rymond and Gorges Poulet,5 have little or nothing to
say aboutjuli and have to emphasize passages from the Rr at
the near total expense of the rest of the work. This could, of course,
be entirely legitimate and it is in accordance, moreover, with the
impact of Rousseau on a literar lineage that includes prestigious
names: Marcel Ryond mentions Maine de Biran, Senancou,
Chateaubriand, Nodier, Neral, Maurice de Guerin, Amiel,
Baudelaire, Rmbaud, and also Gide, Proust, and Rmuz;6 speakng
of "la conscience de soi comme hantise" he refers to a tradition
beginning "more precisely with Russeau" which includes "within
Rmanticism, sbolism, and existentialism, Baudelaire, Amiel,
Kerkegaard, Nietzsche, Mallarme, Valer, Kf. "7 The historical
investment in this interpretation of Rousseau is considerable, and
one of the more intriguing possibilities inherent in a rereading of
Juli is a parallel rereading of texts assumed to blong to the
genealogical line that is said to start with Rousseau. The estence of
historical "lines" may well be the frst casualt of such a reading,
which goes a long way in explaining why it is being resisted.
Serious attempts to come to terms with the structue and dic
tion of Juli have alwas tended towards
_
_ 1i-p<lar,
p
seuc
.i

!c
C

i
n
.
S
the main issue being the defnition of the poles
that set up the tension of the textual feld. Schiller referred to them as
sensitivit [Empjnun] and intellect [Denkkraf] or, in terms of
genre (implicitly) as idyll and eleg,1 a more productive opposition
since it is based, in his terminolog, on the absence or the presence of
5. Marcel RyondJeanaqu Rus au, L qu d si e l r (Paris,
196). Numerous references to Russeau are sattered thougout the wtin of
Gores Poulet; se mor spcifcally the chapters on Rousu in Et sur I tmp
h
umin (Paris, 199) and in L mamrh du cl (Paris, 191) .
6. Ryond,Jeanaqu Rus au, L qu d soi e l r, p. 1 5.
7. Ibid. , p. 193.
8. Friedrch Shiler, Ub ni und sntmntalihe Dihtng, i n Shil1s
Werke, Nationaiausgb edited b Bnno von Wiese and Helmut Koopman (Weimar,
192) , 20:451.
ALL EGORY (JULIE) 191
a referential moment . 9 Dialectizing Marcel Ryond,j Starobinski 1
reads the novel as a tension between immediac (transparence) and
mediation (obstacle). But whatever name is gven to the polarities.
s
g
enerally admitted that the dialectical progression fails. The ten
sion between immediac and mediation allows th
e

ro
raton- of
the experience of nature wth that of an individual consciousness that
overcomes its alienation by an act of love: "la transparence des
cours restitue a la nature l' eclat et l'intensite qU'elle avait prdus"
("the transparenc of the heart restores to nature the brightness and
intensit that it had 10st
, ,)
.
" 1 0
It also makes it possible to pass from
indivdual passions such as love, to the collective and social dimen
sions of the state. Here difculties begin to arise, for the political and
economic theor of Clarens proves to be something of an embar
rassment to anyone who attributes to Russeau the belief that a
political order is conceivable only if it allows for an unmediated
presence of consciousness to itself. Some assimilate the political
model of Clarens to a utopia 11 or denounce it as mere to
talitarianism; 1 2 others, like Starobinski, tr valiantly to rescue what
can be rescued, but are forced to conclude that ':ean-jacques ap
pears to us as a restless soul that falls prey to the power of ambiva
lences, and not as a thinker who posits thesis and antithesis."13 The
ambivalence is at its most evident in the passage from political to
religious language, in the fnal confct between julie's faith and
Wolmar's atheism, which appears to be a gesture of evasion before
the unresolved contradictions of the political world: "To the earthly
well-being that could have been a 'reasonable' ending of the Nouvl l
Heloie, Russeau opposes an alternate conclusion that is religious in
nature." 14 Within this religious consciousness, the same incapacit to
reach a genuine synthesis persists: on the one hand, like the Vicaire
Savoyard (read superfcially), julie seems to advocate a theophany, a
natural religion; on the other hand, an unmediated encounter with
Gd is still being promised in a realm that lies beyond death: "follow
ing the laws of an almost Manichean dualism that radically sepa-
9. Ibid. , 20:4449 and note.
10. Starobinsk,jean1aqu Ru au, L tranprnce e l 'btal, p. 105.
1 1 . For example, Judith N. Shklar, Men an Citizn, A Stu ofRus eau's
Sil Thr (London, 1969) .
12. For example, Lester CrockerJ. j. Ru au ( New York, 196) .
13. Starobinski , jean1aqus Ru au, L tnparn et l'obtal, p. 142.
14. Iid. , p. 140.
192
ROUS SEAU
rates spirit from matter, death causes t he abolition of all obstacles,
the disappearance of all mediations. "1 5 The dialectics of love and of
plitics are fnally superseded by a religious experience that is no
longer dialectical in any sense and that simply obliterates the entire
exprience that precedes it: "Rousseau . . . ends his novl in a man
ner equivalent to a choice. Betwee!._1QluJ. ir erative of the
(mmuni! and that ofpers
QLg.tio.o .. bJlas chosen the latter.
Julie's dea s! fes thil ..dice. "1 6 The literar consequence of this
decision takes the form of a return to a confessional, Augustinian
mode. In this reading tooJuli ultimately appears as only a momen
tar aberration left far behind by a spiritual experience that tran
scends it. Or, if one asserts that Russeau always remained tempted
by the ethics of political reform and by the seductions of the novel ,
then he failed to make up his mind although he was able to articu
late clearly t_he _it.Jr tl. h!fe. Perhaps the failure of the
dialectic is not the failure ofulibut the unavoidable consequence of
positing an antithetical model where none exists. Wich compels us,
however, to discover relationships which, in Wordsworth's terms,
would have "another and a fner connection than that of contrast. " 1 7
In the ver passage in which Julie speaks of an encounter wth
Gd, the encounter is not described as a transparenc but by means
of a metaphor, the curiously unreadable metaphor of reading which
one never seems to want to read. The communication does not occu
in the form ofa perception ("The eternal Being . . . spea neither to
the eyes nor to the ears, but to the heart") , but the contact which can
be called unmediate< because it does not involve a sense perception
occurs as a "reading": it is "an un mediated communication, similar
to the one by which G rea our thoughts already in this life, and
by which we wll , in tum, ra his thoughts in the afterlife, since we
will see him face to face" (2: 728 my italics) . A note draws further
attention to the verb "to read": "This seems to be ver well put . For
what does it mean to see G but to read in the supreme intellect?" l !
15. Ibid. , p. 145.
16. Ibid. , p. 145.
17. William Wordsworth, "ESsy upn Epitaphs," in Th Pe Work of
Willim Wordorth, ed. W. J. B. Owen and Jane Worthington Smyser (Oxford,
1974) , 2: 53.
18. A further gloss appars a few paragaphs later in the sme letter: Julie
speculates as to whether or not she wll "see" afer death those who were dear to her
on earth and Russeau comments: "II est aise de comprendre que par Ie mot voir elle
"LLEGORY (JULIE)
19
The action which Starobinsk rightly considers as the crux of the
entire text (although Julie seems to spea of it with less solemnity
than the commentator) 19 is represented by Russeau as the act of
readig. All the thematic problems of the work, the relationship
between love, ethics, political society, religous experience, and their
respective hierarchies, depend on the understanding of a term of
which the meaning, for Rousseau, is by no means transparent. What
does the Nouvll He10fe have to tell us about the problema tics of
reading?
People read a lot in this book, for there can be no better way to
thematize the ever-present necessity of reading than the epistolar
novel. Unlik Lados's letters in the Liion dngeus, which are
as directly efective as bullets, the letters ofthe Nouvelle He10fe rarely
set out to accomplish anything specifc beyond their ow reading;
apparent deviations from this norm would turn out , at more careful
consideration, to b hardly exceptions at all . 20 Rousseau's text does
en tend un pu acte de l'entendement, sembIable a celui pr lequel Dieu nous voit et
par lequel nous verrons Dieu. " The rational act of understanding (entrnt or
Vemunf) here called "voir" is conceived as "lire. "
19. "Au reste, ajouta-t-elle en regrdant I ministre d'un air assez gai , si je ne
me trmp, un jour ou deu d'ereur sernt bientot passe. Dans pu j'en sauai
la-dessus plus que vous-meme." (2: 79) or, wth reference to the possible immortal
ity of the soul : "C'est une folie, soi t, ris elIe est douce . . . " (2: 695, note c) .
20. It is obous that the narrative does not exploit various possibilities that
would equate the letters wth acts rather than with medi tations or discourses. There
is, for example little fetishim in which the letters act as a mere substitute for the
body of the belovd; in fetishistic senes, the olect is julie's painted prtrait or her
house, not her letters. The facticity of the letters, when it is referd to, is not as a
substitutive presence, but in a curiously li teral way, as when a note is added about
the frequenc of mail deliveries (2: 71) . The discursive letters (from the VaIais, from
Paris, on music, etc. ) ar judged b julie to be inopprtune rather than seductive,
and quite superuous from the point of view of fostering the practical afairs of the
two lover. When a fetishistic substitution occus (" Baise cette lettre et saute de joie
pour la nouvelle quej vais t'apprendre . . . " [2: 1 11]) , it is precisely when something
more tangible than a let ter is being promised and ver practical arrangements are
being made. A highly dramatic exception seems to be letter 25, Part I, in which
Saint-Preu threatens to kll himslf ("L roche est escarpee, l'eau est profonde et je
suis au desespoi . . ." [2:93]) . This appars to lead to julie's surender and to make
the let ter into an efectiv act of sduction. But the immediac of this cause and
efect relationship is more illusor than actual. Other
e
lements interene and more
time elapses. The point is that julie gives in to pity rather than to the direct expres
sion of desire and Let ter 25 can be considered as part of a development on the theme
of pity rather than as a sductive strateg. One could say, of course, that julie's
19
ROUSSEAU
not exploit the narrat ive possibilities of the let ters as "actants," as
direct plot-agents. The rather appar to be reflective and retrospc
tive musings , interpretations of events rather than being themselves
the events. This is, in part , why the novel has little difcult in
representing the facticit of reading but faces some awkward mo
ments when it comes to wrting.
2
1 Hence also the Ciceronian, de
clamatO stle destined to be read aloud and heard rather than
visualized-probably the main obstacle to the enjoyent ofthe novel
by the contemporar reader. Not that the act of reading is innocent ,
far from it . It is the starting pint of all evil . "The woman who, in
spite of the title, will dare to read one single page, is a lost woman
. . . " (2:6) , is a sentiment echoed by Julie as the reader of Saint
Peux's letters: ". . . you wrote. Instead of throwing your frst letter
into the fre or takng it to my mother, I dared to open i t. This was
my crime and all the rest followed. I tried to force myself not to
answer these nefarious letters which I could not prevent myself from
reading" (2:32) .
The evil ofthe letters can be too easily attributed to their literar
mediation, to the desire they convey in the guise of fctions. As we
k(W from the Ea
y
, Russeau claims that seduction could be much
more efectively performed by mimicr and by gesture than by writ
ing; part of the realistic oddit, bordering on the ludicrous, of the
novel is that the letters are so didactic in tone, and the distance
between Saint-Preux and Valmont so hard to bridge. The entire
reception ofuli goes in a diferent direction, but Laclos's, Hazlitt's,
or Stendhal' s use of the novel almost parodies the obvious misreading
predicted by Russeau. The letters are no invi tation to a shared erotic
or passionate exprience and to read la Nouvell Heloie the way
Paolo and Francesca read Lancelot only results in its dismissal as a
bourgeois version of the Tristan myth.
22
Dspite the chivalric conno
tations of his name, Saint-Peux's literar archete is closer to
Abelard than to Lancelot or to Tristan. What Julie and Saint-Peux
invtation to Si nt-Preux, in Ltter 4 of Part I, tojoin her and Wolmar at Clarens is a
d
irect action, but this is certainly rather a mechanical need of the plot and the letter
consists of only two and a half lines.
21. As in letter 15, Part I (2: 147) , when Saint-Preu i about to enter Julie's
bdrom and interrpts his exalted anticipati ons wi th the remark "Quel bonheu
d
'avoir trouve de l'encre et du papier! "
22. Dnis de Rugemont , L'amur et [ 'occint (Pans, 1939) .
ALLEGORY (JULI 195
will read together is austere fare, and it includes "neither poets, nor
love stories, contrar to the usual readings destined to Uulie's] sex"
(2: 61 ). 23 Their relationship is not primarily characterized as a desir
mediated by literar substitutes; there is little reason not to tae
Russeau at his word when he has Saint-Preux say "Wat would we
learn about love in book? Ah, Julie, our heart has more to tell us
than they can, and the imitated language of book is cold for anyone
who is himself passionate" (2: 61) . The temptations emanating from
literar inscriptions, as in the lines from Petrarch and Tasso which
Saint-Preux engraved on the rock at Meillerie (2: 519) are genuine
dangers and the text tries, at the risk of heroic boredom, to avoid
having a similar efect . The abundant presence of literar antece
dents much in evidence throughout the novel, in direct quotation as
well as by allusion, are never merely a quiotic mystifca
-
tion that
would imply a simple displacement of a desire upon a text . When
such patterns occur, the are only a minor version of a more inclusive
structure. Intertextuality in l Nouvell Heli is more than just
"literar" in its complications. The danger of reading is a far
reaching and invidious threat that no conversion, however radical ,
could ever hope to remove.
The best place in the text ofuli to enrich one's understanding
of "reading is without doubt the second Preface, sometimes referred
to as "Dialogue on the novel" and staging a confrontation btween
author and reader in the conventional form of an apologia (2 : 12-30) .
Despite its largely traditional terminolog, this brief text has little in
common with the habitual eighteenth-centur discussions of the rel
ative merits of fction as compared to histor. 24 It therefore demands
23. It is true that the list contains Petrarch, Tass, Metastasis, and ' the French
tragedians. Russeau considers Petrarch and the Italian poets in the wae of the
Augstinian rather than the chivalric tradi tion. In the case of Rcine, one may
assume that he is thinkng of Esther rather than of Phedre. The ambivalence of the
literar valori ztions is part of the wider ambivalence of all systems grounded in
metaphor.
24. See Werner Kaus. "Zur Franzosischen Romant heorie des 18. jahrhun
derts" in Nachahmung un Illusion, Poetik und Hermeneutik 1, H. R. jauss, ed.
( Munich, 1964) , for a brief bibl iography of the question. Some of t he quotat ions
from lesser kown aut hors ' are ver dose to the distinction from which Rousseau
starts out . In the sme volume, the discussion of Kaus's paper is i nformat ive.
There woul d, of course, be numerous English exmples , perhaps well summari zed
in the openi ng chapter of Bok I I I of Fi eldi ng'sJosph Andrws ( "Matter prefat or
in praise of biogaphy") .
19
ROUSSEAU
a reading i n its ow right, much asjuli itselffails t o conform t o the
norms and conventions of eighteenth-centur fction.
The central question around which the imaginar debate of the
preface circles is not that of verisimilitude (granted by both inter
locutors to be nonexistent injuli, in realistic detail as well as in the
general conception) , but that of the text's referential status. Does the
moel for the main characters in the narrative exist outside the
language of fction or not? "Is this corespondence real or is it a
fction?" The device (presenting a fction as if it were a histor) is
common enough and coincides wth the emergence of the novel as a
separate genre, yet Russeau's treatment of it goes a long way in
explaining its almost obsessional recurence in the histor of the
novel . 25
The dialogue starts out from what appears to be a classical
antithesis: a narative text can be either the "portrait" of an extra
textual , particular referent, or a "tableau. " The "tableau" does not
have a specifc referent and is therefore a fction ( "tableau d'imagina
tion," (2: 9]) . Common sense tells us thatjulie is a tableau and Rus
seau states as much, in no uncertain terms, in the actual , frst Pref
ace. 26 It may, of course, be a displaced portrait transfered into a
fction, but this, for the moment , is not the issue. We were never
supposed to take literall the assertion of the title page that the
letters of the two lovers, "inhabitants of a little town at the foothills
of the Alps ," have been "collected and published" (and not written)
by J. J. Rousseau. This does not mean however that the reverse ofthis
proposition, namely that Russeau is the author of these letters in the
full sense of t he term, is simply true wthout any reserations. All
assertions to the contrar, a "question" seems to re
m
ain and to
demand exploration, as if the either/or choice between "portrait"
and "tableau" were perhaps not as mutually exclusive as might seem
to be the case.
If juli is not simply a "portrait ," a mimesis of an action or a
25. The closest antecedent for Juli is, of course, Rchardson's Clrs, but
Russeau refers most frequently to Montesquieu's I Temple d Gni, which is
mentioned at least once in L nouvlle Helie (2: 1 13) and reappars in the Fourh
Rr.
26. As distinguished from the Preface dialogUee. "Quant a la vrte des faits, je
declare qu'ayant ete plusieurs fois dans Ie pays des deu amants , je n'
y
ai jamais ouj
parler du Baron d'Etange ni de sa flle, ni de M. d'Orbe, ni de Milord Edward
Bomston, ni de M. de Wolmar" (2: 5).
ALLEGORY (JULIE) 197
person, what then does it mean for a text to be a tableau, a fction?
For the enlightened reader that N. seems to be (call him Marmontel
or almost any witer on fction at the time) the way back to referen
tialit and to meaning or truth is eas enough: fction and the trth
are reconciled through the concept of man as universal : "ever
human fgure must possess the common traits shared by al men,
otherse the fction [tableau] is worthless"; " . . . in the fctions
[tableaux] of humanit everone must be able to recognize Man"
(2: 12) . That the idea of man as a well-established, stable concept has
to be rejected by Rousseau is predictable enough, since the Scon
Diourse and the Esay chose the ver word "man" as the target of
their epistemological critique of concepts in general . Behind the reas
suring term lurk an unknown, unpredictable, and unreliable mon
ster or "giant . " N. 's protestations bring the inferences of Russeau's
anthropolog into the open: " . . . [in the absence of the universal]
unheard of monsters, giants, pygees, chimeras of all sorts, any
thing could spcifcally b admitted within nature: everhing woud
be disfgured, we would no longer have a common model . . ."
(2:21) . Without takng away the terror, the same feeling ofunprdict
abilit can just as well b phrased in positive ters, as an assertion
of freedom, of infnite possibilities and renewals, as in the quotation
from the Scil Contat that Holderlin was to single out as a motto
for his earl "Hymn to Humanit : "The limits of the possible in
spiritual matters are less narrow than we assume. Our weaesses,
our vices, our prejudices are the cause of our confnement . Lowly
souls fail to believe in great men; vile slaves smile scornfully at the
mention of the word freedom. " (3:425) . The same positive tone can
be heard in the Second Preface: "Who would dare to assign precise
limits to nature and assert : This is how far man can go, and not
further?" (2:20) , or, in the reversed value-pattern but still wi thin the
same metaphor: "0 Philosophy! How you take pains to make hearts
narrow and man small! " (2:28) .
The pathos of these statements, regardless of whether the are
expressions of terror or assertions of prophetic exaltation, stems from
the referential indeterminac of the metaphor "man. " The an
thropological "tableau" is indee a fction, bewildered by its ow
suspended meaning. It depicts human passions (fear, pit, love,
freedom) but these passions all have, by defnition, the self-deceiving
structure, familiar to us from the Scon Dicourse, that forces the
narrative of their deconstruction t o unfold. In the case ofuli, the
:9
ROUSSEAU
passion happens to b lov, the traditional topic of romance, but one
could just as well imagne novels of pure fear or, like Proust' s, of pure
suspicion. On the thematic level, it is not the prevalence of love
rather than fear that sets apart Juli from the Dicours, since the
determining strcture is that of passion which both have in com
mon.
Like "man," "love" is a fgure that disfgues, a metaphor that
confers the illusion of proper meaning to a suspended, open seman
tic structre. In the naively referential langage of the afections, this
makes love into the forever-repeated chimera, the monster ofits ow
aberration, always oriented toward the future ofits repetition, since
the undoing ofthe illusion only sharpens the uncertaint that created
the ilusion in the frst place. In this same afective language, the
referential error is called desire, and the voice of this desire can be
heard througout Rousseau's writings: "Such is the nothingness of
my chimeras that, if all my dreams had turned into realities, . ! would
still remain unsatisfed. I would have kept on dreaming, imagining,
desiring. I found in myself an unexplainable void that nothing could
have flled, a longng of the heart towards another degree of fulfll
ment of which I could not conceive but of which I neertheless felt
the attraction" (Ltr t Malhrbe, Ouve complt, 1: 1140) .
The Second Peface says the same thing in slightly more technical
terms by establishing the link wth fgural diction: "Love is a mere
illusion: it fashions, so to speak, another Universe for itself; it sur
rounds i tself with objects that do not exist or that have received their
being from love alone; and since it states all its feelings by means of
images, its language is always fgural" (2:46) .
However evanescent the referent of the passion may have be
come ("Ie neant de mes chimeres," "Ie neant des choses humaines"
Juli, 2:693) , it is clear that once the fgurality of the langage of
passion has been established ("son langage est toujour fgure") we
return in fact to a referential moel . The unproblematic fgrali t of
the metaphor restores its proper meaning, albeit in the form of a
negating power that prevents any specifc meaning from coming into
being. The ver pathos of the desire (regardless of whether it is
valorized positively or negatively) indicates that the presence of de
sire replaces the absence of identit and that , the more the text
denies the actual exstence of a referent, real or ideal , and the more
fantastically fctional it becomes, the more it becomes the repre
sentation of its ow pathos. Pathos is hyostatized as a blind power
ALLEGORY (JULIE) 19
or mere "puissnce de vouloir," but it stabilizes the semantics of the
fgure by making it "mean" the pathos of its undoing. In the text of
the Second Preface, the speaker referred to as R. (standing presum
ably for Russeau) , by unsettling the metaphor "man," shifs from
anthropological generality to pure pathos, but the fgrality of the
language of love implies that pathos is i tself no longer a fgure but a
substance. In the terminolog of the text , the "tableau" has become a
"portrait" afer all , not the portrait of universal man but of the
deconstrctive passion of a subject. R. 's snsible interlocutor under
stands this ver well and knows that , as soon as the return to mimet
ic representation is in fact granted and claimed to be internalized,
the impersonal desire is again susceptible of being represented by a
subject: "R: So we will fnd men, in books, only as they wish to reveal
themselves? N: The author as he wishes to reveal himself; the
characters he describes such as they are . . ." (2: 14) . R. started out
by deconstrcting the referential system based on the metaphor
"man" but has substituted for it a new referential system based on
the pathos of a temporal predicament in which man's self-defnition
is forever deferred. The polarity between "portrait" and "tableau"
does not engender extra-textual referents. The inside is always al
ready an outside.
In the process of this discover, however, the orginal system
undergoes some transformations. At the onset of the text , "tableau"
and "portrait" were associated with author and editor respectively: if
the work was imagnar, then Russeau had to be the author; if it
were to be an actual collection of letters, the portrait or copy of a
witten text, then Russeau was merely the editor. We then move on
to a work that unsettles its ow referential status but fnd that such a
work can be read as the "portrait" of its ow negative gesture. It
follows that , if the work indeed represents objects "qui ne sont
point ," then it is the "portrait" of a suqject's initiation to this knowl
edge. But only this subject can be the author of the text , since Julie,
the emblem of love, is par excellence the object that does not exist .
Rousseau i s then the author of what turns out t o be the portrait of an
impossible "tableau. " The original pairing of author with "tableau"
has now been reversed, and instead of being paired with editor,
"portrait" is now paired with author. As such, unlike the literal
portrait of the real julie, it becomes again comprehensible and "in
teresting." For the original pairng was self-defeating: "I understand
you' , says R. 'If these letters are portraits, they are of no interest ; if
ROUSSEAU
they are fctions, the imitate badly''' (2: 10) . This original impasse is
broken in deconstrctive narratives of which the Scon Discours
and the Es y are examples: the imitation is now, in epistemological
terms, a "good" imitation for it is free of any trace of distortion or
wishful mystifcation; on the other hand, it is "interesting" since it
portrays a pathos in which all can share. Contrar to received opin
ion, deconstructive discourses are suspiciously text-productive.
It is not only possible but necessr to read Julie in this way, as
putting in question the referential possibili t of "love" and as reveal
ing its fgural status. Such a reading would difer from the available
interretations of the novel but would not be essentially diferent
from our reading of the Scon Discourse and of the Esay
: both are
deconstructive narratives aimed at metaphorical seductions. From a
rhetorical point of view, nothing would distinguish the discursive
language of the earlier texts from the language of the novel . Such a
reading is a necessar part of the novel's interpretation, which has to
start out by undoing the simply antithetical relationship between
referent and fgure. This does not mean however that it can stop
there.
Rousseau himself, at any rate, does not , in the Second Preface,
allow the reading to come to rest. The question of authorship never
receives a satisfactor answer, although it would seem to be a settled
matter. N. keeps pressing R. to afrm or deny his authorship but R.
keeps refusing, not for reasons of prudence, modesty or shame, but
in the name oftrth. "N. : If! keep asking you whether or not you are
the author of these letters, why do you avoid my question? / R: For
the ver reason that I do not wsh to tell a lie. / N. : But you also refuse
to tell the truth? / R: Declaring one's wish to keep truth unsaid is still
a way to pay tribute to it . . . " (2:28) . "Taire la verite" does not mean
here to conceal something one knows, but not to proclaim kown
something one is unable to ascertain: "Who could decide whether or
not I am caugt in the same doubt in which you fnd yourself:
Whether all this myster and evasion is not a feint in order to hide
my own ignorance of what you are tring to discover?" (2:29) . Wat
can it mean, in this context , for the author of a text to claim that he
doesn't know whether or not he is its author? We speak perhaps too
easily nowadays of the impersonalit of witing, of wting as an
intransitive verb, "disparition elocutoire qui laisse l'initiative aux
mots" (Mallarme) . Are these the terms in which we are to under
stand Rousseau's statement?
ALLEGORY (JULIE) 201
Part of the difcult stems from a naive distinction between
"witing" and "reading," from an exclusive concentration on author
ship at the exclusion of the reader. Unlike the Sn Diours and
the Eay, which deal with the orignation of langage and con
sequently wth "wting" the Scond Peface to Juli deals wi th read
ing in its relationship to witing. If to read is to understand
writing-and we are deliberately leaving aside, for the moment , the
performat ive function of langage-then it presupposes a possible
kowledge of the rhetorical status of what has been witten. To
understand primarily means to determine the referential mode of a
text and we tend to take for ganted that this can be done. We
assume that a referential discouse can be understood by whoever is
competent to handle the lexcological and gammatical code of a
langage. Neither are we helpless when confronted with fgues of
speech: as long as we can distinguish between literal and fgural
meaning, we can translate the fgre back to its proper referent . We
do not usually assume, for example, that someone sufers from hal
lucinations merely because he says that a table has four legs; the
context of common usage separates the fgral meaning of the
catachresis (which, in this case, leads to the referent) from its literal
denotation (which, in this case, is fgal) . Even if, as is ofen said to
be the case for poetic language, the fgre is polysemous and engen
ders several meanings, some of which may een b contradictor to
each other, the large subdivision beteen literal and fgral still
prevails. Any reading always involves a choice beteen signifcation
and smbolization, and this choice can be made only if one postu
lates the possibility of distingishing the literal from the fgral . This
decision is not arbitrar, since it is based on a variet of textual and
contextual factors ( grammar, lexcolog, tradition, usage, tone, de
clarative statement , diacritical mark, etc. ) . But the necessity of mak
ing such a decision cannot be avoided or the entire order of discourse
would collapse. The situation implies that fgural discourse is always
understood in contradistinction to a form of discourse that would
not b fgural ; it postulates, in other words, the possibilit of refer en
tial meaning as the tl of all langage. It would be quite foolish to
assume that one can lightheartedly move away from the constraint
of referential meaning.
The critical thrst of Russeau's theor of language in the Sc
on Diours and in the Esay undermines this model . In these
texts, the discussion of denomination as the primal linguistic func-
28
ROUSSEAU
ton i n fact puts i n question the status of referential language. I t
becomes an aberant trope that conceals the radical fgralit of
langage behind the illusion that it can properly mean. As a rsult,
the assumption of readabilit, which is itself constitutive of lan
gage, cannot only no longer b taken for granted but is found to be
aberrant. There can be no writing wthout reading, but all reading
ae in error because the assume their own readabilit. Everhing
wtten has to be read and ever readng is susceptible of logical
verfcation, but the logic that establishes the need for verifcation is
itself unverfable and therefore unfounded in its claim to trth.
In the Second Prtface to juli, N. , as a reader, is dependent on
the possibility of reference and represents this need metaphorically
by the assumption that the author holds the ke to the referential
status of his langage. Hence his tireless questioning of R.'s author
ship, in itself a correct representation of the necessarily naive com
ponent included in any act of reading, regardless of its level of com
petence. This elusive author is not initially a subject but the
metaphor for readabilit in general . Since he interenes only to the
extent that he is supposed to control the rhetorical mode of the text,
he becomes the metaphor of a will or of a subject. Unlike N. , Rous
seau is supposed to kow whether the text of his novel was merely
copied (or quoted?) from a previous document or whether he in
vented it as his own creation. Athough at this particular moment in
the dialogue (2: 12) N. expresses a preference for the frst alternative,
he could accommodate himself to both possibilities. Asked whether
he can respond to the pathos of the text , he replies: "I can conceive of
this efect wth regard to you. If you are the author, the impact is
easy to understand. If you are not, I can still conceive of it . . . "
(2: 18) . What he could not tolerate, however, is the impossibilit of
distinguishing between the alternatives. This would leave him dan
ging in an intolerable semantic irresolution. It would b worse than
madness: the mere confusion of fction with reality, as in the case of
Don Quiote, is mild and curable compared to this radical dyslexia. 27
27. "Voulant etre ce qu'on n'et pas, on paIent a se croire autre chose que ce
qu'on est , et voila comment on devent fou. En montrant sans cesse a ceu qui les
lisent les pretendus chanes d'un etat qui n'est pas Ie leur, [les romans] les seduisent,
ils leur font prendre leur etat en dedain, et en faire un echange imaginaire contre
celui qu'on leur fai t aimer. " (2:21) . This is a simple fgral exchange in which t he
t
w specular poles, "leur etat" and "un etat qui n'et pas Ie leur," are dearly to b
distinguished. The madness can be considerd the madness of another which does
not threaten the sanit of t he reader.
ALLEGORY (JULIE) 20
R.'s denial ofthe kowledge that he is the author of his ow text
leaves N. in a predicament that his imagination cannot even begin to
gasp. He would much rather assume that R. is mystifing him
deliberately by wthholding information that must be in his posses
sion. The relationship between author and reader would then be one
of simple deceit . The author is a liar, an unreliable narrator open to
moral censure21 or suspect , at best , of playing a frivolous game of
hide and seek. The novel would be a riddle rather than an eniga,
wth a defnite answer know from the beginning and atifcially
encrted, lik the missing body in a clumsy myster stor. The text
would be generated by the mere deferment of a kow secret. It is
painfully clear that this is not the strcture of the Nouvell He1oie, a
novel not distinguished by dramatic suspense.29 The only suspense
would stem from Rousseau's capricious withholding of the key to his
roman i clef
Taken literally, Rousseau's assertion that he does not know
whether he or his fctional characters wote the letters that make up
Juli makes little sense. The situation changes when we realize that
R. is merely the metaphor for a textual propert (readabilit) .
Further inferences then become apparent , for example that R. is
similar to N. in his inability to read }uli and that it is impossible to
distinguish between reader and author in terms of epistemological
certainty. It follows that we can reverse the priority which makes us
think of reading as the natural consequence of wrting. It now ap
pears that witing can just as well be considered the lingistic cor
relative of the inability to read. We write in order to forget our
forekowledge of the total opacity of words and things or, perhaps
worse, because we do not kow whether things have or do not have
to be understood.
In the Second Preface, we come closest to being draw into the
wake of this whirlpool in passing from the terminolog of reference
and fgure (portrait and tableau) to that of textualit. In a search
parallel to his quest for authorship, N. wants to fnd a statement
within the text that establishes the margn between text and external
referent, that clearly marks of an intra-textual from an extra-textual
2. Se Waye Bt h, Th Rr ofFin (Chicago, 1961) , Chapters 12
and 13.
29. "N. Quant a !'interet , i l est pour tout I e monde, il est nul. Pas une muvaise
acton; pas un mechant homme qui fas craindre pour les bons. D eenements si
naturels, si simples qu'ils Ie snt trop: rien d'inopine; pint de coups de Theatre.
Tout est prevu longtemps d'avance; tout arrive comme i l est prevu . . . " (2: 13) .
21
ROUSSEAU
feld. He thin t o have found i t i n the epigraph on the title page, a
passage from Petrarch which, in Rousseau's own translation, reads
as follows: "L monde la posseda sans la connaitra, et moi je l'ai
connue, je reste ici-bas a la pleurer" (2: 1338) . "Don't you see," asks
N. , "that your epigraph gives it all away?", assuming that the author
confesses in this way the exstence of the live model. The authority of
the quotation is, of course, anyhing but decisive: it is highly ambiva
lent in itself; it is not Russeau's statement but is borrowed from a
complex context ; it is not even Petrarch's statement , since Petrarch
borrows it freely from John the Evangelist where it refers to Gd
as Logos, etc. To all these possibilities of doubt , R. adds one less likely
to come to mind: "for who can know whether I found this epigraph
in the manuscript or whether it is I who put it there?" (2:29) . Even if
Rousseau had merely copied the letters, this would in no way estab
lish their referentiality, since they might have been witten by some
one who, as the use of epigraphs shows, was just as much in need of
reassurance as to the status of his text as R. and N. admit to being.
The author of the letters may not have acted, copied, or portrayed
but merely quoted. And it is impossible to say where quotation ends
and "truth" begins, ifby truth we understand the possibility of refer
ential verifcation. The ver statement by which we assert that the
narrative is rooted in reality can be an unreliable quotation; the ver
document , the manuscript , produced in evidence may point back,
not to an actual event , but to an endless chain of quotations reaching
as far back as the ultimate transcendental signifed God, none of
which can lay claim to referential authority.
The Second Preface toJuli thus link a de constructive theor of
reading with a new sense of textuality. The innumerable writings
that dominate our lives are made intelligible by a preordained
agreement as to their referential authority; this ageement however
is merely contractual, never constitutive. It can be broken at all times
and ever piece of writing can be questioned as to its rhetorical
mode, just asJuli is being questioned in the Preface. Wenever this
happens, what originally appeared to be a document or an instru
ment becomes a text and, as a consequence, its readability is put in
question. The questioning points back to earlier texts and engenders,
in its turn, other texts which claim (and fail) to close of the textual
feld. For each of these statements can in its turn become a text , just
as the citation from Petrarch or Russeau's assertion that the letters
were "collected and published" by him can b made into texts-not
ALLEGORY (JULlE) 205
by simply claiming that they are lies whose opposites could be true,
but by revealing their dependence on a referential agreement that
uncritically took their truth or falsehood for ganted. The same
applies, of course, to the text of the Preface with regard to the main
text ofuli: rarely has a preface been less able to shed light on the
meaning of the text it introuces, to the point of thematizing this
impotence into the kowledge of an ignorance which the main text
will , in its turn, have to challenge. We can no longer be certain, at
this point, whether the preface was written for the main text or the
main text for the preface.
The rhetorical mode of such structures can no longer be sum
marized by the single term of metaphor or of any substitutive trope
or fge in general , although the deconstruction of metaphorical
fgres remains a necessar moment in their production. They take
into account the fact that the resulting naratives can be folded back
upon themselves and become self-referential. By rin, for reasons
of epistemological rigor, to confrm the authority, though not the
necessity, of this juxtaposition, Russeau unsettles the metaphor of
reading as deconstructive narrative and replaces it by a more com
plex structure. The paradigm for all texts consists of a fgure (or a
system offgures) and its deconstruction. But since this model cannot
be closed of by a fnal reading, it engenders, in its tun, a
supplementar fgural suprposition which narates the un
readabilit of the prior naration. A distingished from primar
deconstructive naratives centered on fgures and ultimately always
on metaphor, we can call such narratives to the second (or the third)
degee allegor. Allegorical narratives tell the stor of the failure to
read whereas tropological narratives, such as the Son Dicoure,
tell the stor of the failure to denominate. The diference is only a
diference of degree and the allegor does not erase the fgue. A
legories are always allegories of metaphor and, as such, they are
always allegories of the impossibility of reading-a sentence in
which the genitive "of' has itself to b "read" as a metaphor.
In the text of the Second Preface, the point at which the allegori
cal mode asserts itselfis precisely when R admits the impossibilit of
reading his ow text and thus relinquishes his power over it . The
statement undoes bth the intelligbilit and the seductiveness that
the fction owed to its negative rigor. The admission therefore occurs
againt the inherent logc which animated the development of the
narrative, and disarticulates it in a way that seems pererse, just as
20
ROUSSEAU
Russeau's discursiveness, or Julie's preaching, has seemed to all
critics the p
rerse spoiling of a fne subject . The reversl seems
opposed to the best interests of the narrator. It has to be thematized
as a sacrifce, a renunciation that implies a shif in valorization.
Before the reversl , the narrative occurs within a system governed by
polarities of truth and falsehod that move parallel with the text they
generate. Far from interfering wth each other, the value system and
the narrative promote each other's elaboration; hence the relative
ease of the narrative pattern in the SoM Discourse despite (indeed
because of its fgural complications, or of the stor of passion in the
frst part ofulie which is sid to be "like a live source that flows
forever and that never runs dr (2: 15) . But in the allegor of un
readabilit, the imperatives oftruth and falsehod oppose the narra
tive syntax and manifest themselves at its expense. The concatena
tion of the categories of truth and falsehood with the values of right
and wrong is disrupted, afecting the economy of the narration in
decisive ways. We can call this shi f in economy ethical, since it
indeed involves a displacement from pathos to etho. Alegories are
always ethical , the term ethical designating the structural interfer
ence oftwo distinct value systems. In this sense, ethics has nothing to
do with the wll ( thwarted or free) ofa subject , nor afomri, wth a
relationship between subjects. The ethical categor is imprative
(Le. , a categor rather than a value) to the extent that it is linguistic
and not subjective. Morality is a version of the same language aporia
that gave rise to such concepts as "man" or "love" or "self," and not
the cause or the consequence of such concepts. The passage to an
ethical tonalit dos not result from a transcendental imprative but
is the referential (and therefore unreliable) version of a linguistic
confusion. Ethics (or, one should say, ethicity) is a discursive moe
among others.
But the Preface and the main text ofulie are ethical not only in
this wider sense. They are also moralistic in a ver practical way that
frequently borders on the ridiculous but that is nonetheless a neces
sar part of what is most consistent in Russeau's thought. In the
Preface, this tone, all too familiar to readers of Julie and Emil, is
much in evidence in R.'s lengthy considerations on all the good his
book wll be able to do for its readers. The discrepanc between the
persona of Russeau as the critical moralist of rhetorical suspicion
and that of the man of practical wsdom is puzzling. The relation
ship between the epistemologist and the voice of practical reason
ALLEGORY (JULIE) 207
( Russeau himself speak of "morale de pratique") is certainly, like
the relationship between Rousseau and the to fctional speakers R.
and N. in the Preface, a dialogical rather than a monological one.
From the frst, one has to expect a mental attitude that is highly
self-reflective, persistently aware of the discrepancies between the
formal and the semantic properties of language, fully responsive to
the seductive play of the signifer yet war of their powers of seman
tic aberration. It supposes an austere analytical rigor that pursues its
labors regardless of the consequences, the most rigorous gesture of
all being that by which the writer severs himself from the intelligibil
it of his own text . Yet , while holding up this attitude as morally
exemplar, Rousseau nevertheless, in the same breath, discusses its
consequences in ver practical utilitarian terms. The same person
who is vulnerable to reproaches of sophistr and over-subtlet when
he dodges simple signifcation with phrases such as "si elles furent,
elles ne sont plus, " or with hard-to-follow evasions on the status of
epigraphs, also speaks naively about his desire to be useful : "i n order
to make what one wishes to say usefl , one must frst of all be
intelligible to those who have to make use of it" (2:22) ; "if one wants
to be useful , one should be read in the Provinces"; "perhaps hus
bands and wives wll fnd in this volume views that may make their
labors useful" (2:23) , etc. We fnd back the same mixture of epis
temological refnement and utilitaian naivte that is characteristic
of much of Rousseau's writing, especially when the "poetic" aspects
of his discourse have to be reconciled wth considerations on moral
it and on customs. No such problems occur with the theoretical
aspects of his political thought ; they arise primarily with regard to
the wisdom of practical reason. The question is not the intrinsic
merit or absurdit of these pieces of good advice but rather the fact
that they hve t be uttered, despite the structural discrepanc be
tween their intellectual simplicit and the complexit of the consid
erations on which they are predicated.
The heterogeneous texture of Rousseau's allegorical narratives is
less surprising if one keeps in mind that his radical critique of refer
ential meaning neer implied that the referential function of lan
guage could in any way be avoided, bracketed, or reduced to being
just one contingent linguistic property among others, as is postu
lated, for example, in contemporar semiolog which, like all post
Kantian formalisms, could not exist wthout this postulate. Russeau
never allows for a "purely" aesthetic reading in which the referential
.o
ROUSSEAU
determination would remain suspended or be nonexistent. Such a
rading is i nconceivable on the epistemological , premoral level ,
where it would be the mere play of a free signifer, nor does it exist in
allegor, when the undoing of signifcation has taken on ethical di
mensions and when the object of aesthetic contemplation would be
the beautiful philosophical soul . The impossibilit of aesthetic judg
ment is built within Rousseau's linguistic model as an aberrant
fgre. Suspended meaning is not, for him, disinterested play, but
always a threat or a challenge. The loss of faith in the reliabili t of
referential meaning dos not free the language from referential and
tropological coercion, since the assertion of the loss is itself governed
by considerations of truth and falsehood that, as such, are necessarily
referential . Kant's concept of aesthetic freedom is, in Russeau, a
metaphor for the indeterminac of signifcation and can thus never
be the source of any judgment , nor a license to elaborate modes of
judgment that would no longer be dependent on concepts . It is clear
that Schiller's Lters on Athti Eduaton could never have ben
recommended reading for Emile. The concept , in Rousseau, always
retains a referential moment, the supply of diference that the con
cept ackowledges by concealing it. But since the convergence of the
referential and the fgural signifcation can never be established, the
reference can never be a meaning. In Russeau's linguistics there is
room only for "wild" connotation; the loss of denominational control
means that ever connotation has claim to referential authorit but
no statute in which to ground this claim. When Kant, using music
and the ornamenta
arabsque3
0
as his main example was in fact to
ground aesthetic judgment in nonreferentialit, his semiological in
sight was gained at the cost ofa repression that was to make theoret
ical poetics, a branch of applied linguistics, into aesthetics , a branch
of applied psycholog.
The persistence of the referential moment (which is to be dis
tinguished from the noncogitive, performative function of lan
gage) prevents the confnement of allegor to an epistemological
and ethical system of valorization. Since the epistemological media
tion is kown to be unreliable, and since the narrative of this discov
er cannot be lef suspended in the contemplation ofits own aesthet
ic gatifcation, the allegor speak out with the referential efcac of
30. Knt, Kritk dr Urtilkraj, 1 1 and further passages on music, 40, 42
f. , etc.
ALLEGORY (JULIE) 209
a pra. The ethical language of persuasion has to act upon a world
that it no longer considers strctued like a lingistic system but that
consists of a system of ne d. The Nouvll Helf, for example, can
be read as a thematization of the movement from symbolic,
metaphorical meaning in the frst part to a more contractual te of
meaning in the second part , afer Julie's mariage to Wolmar. At the
same time however Rousseau is concered wth the practical efect
of this thematization on the reader. And although what he actually
says about this may b quite silly (silliness being deeply associated
wth reference) , the co-presence of thematic, exortative discourse
wth crtical analytic langage points to an inherent characteristic of
all allegorical modes. The resuting discourse of pras is however not
only devoid of authort (since it is the consequence of an epis
temological abdication) , but it occurs again in the form of a text . The
Second Preface, however practical it may be in its concers, is not
more of an action than the rest of the novel. Reading is a praxis that
thematizes its own thesis abut the impossibilit of thematization
and this makes it unavoidable, though hardy legitimate, for al
legories to be interreted in thematic terms.
By movng, under the gidance of the Second Peface to juli,
from fgal deconstrction, frst to the theoretical and then to the
practical ethical dimension of allegor, we had to reintroduce the
concept of need [esoin] which orignally sered Rousseau as the
means to distinguish the langage-structured discourse of passion
from nonverbal entities. Just as it is impssible to understand the
historical condition of man wthout positing a fctional state of na
ture, and just as it is impossible for a statement not to connote a
referential meaning, it is impossible for a passion not to hyostatize
a hypothetical need from which it would be the supplementar dis
placement. Passions are then conceived as pathological needs, which
is also why the are afectively valorized in terms of pleasure and
pain. The allegor inevitably shifs to a eudaemonic vocabular. In its
more domestic versions, this vocabular generates the mixture of
erotic sweetness and deceit, of "doux modele" (2: 13) wth "acres
baisers" that hangs over much of Rousseau's fctions. He himself
comparedjuli to the "soave licor" (Tasso) that covers up the bitter
ness of the actual statement, and this slightly nauseating favor
catches the quintessential aroma of Rousseau's necessarily "bad"
taste. One can always console oneself from this cloyness wth the
hygenically brisk Scil Contrat.
210
ROUSSEAU
With the reintroduction of needs, the relapse into the seductions
of metaphor is inevitable and the ccle repeats itself Needs reenter
the litera discourse as the aberant proper meaning of metaphors
againt which the allegor constitutes itself. The reintroduction of
the intentional langage of needs into the allegor is not itself inten
tional but the result of a lingistic structure. The entire assumption
of a nonverbal realm governed by needs may well be a speculative
hyothesis that exsts only, to put it in all too intentional terms,for
th sa oflanguage. But the existence of this moment of relapse in
Russeau's allegories frt has to be documented and, by alluding to
its exstence, we have in fact moved away from the Second Preface
into the main text ofuli.
The len
g
thy recapitulative letter (2: 340-65) in which Julie explains to
Saint-Preux the reasons for her mariage can sere as a (fallacious)
synecdoche for the totalit of a text which demands a much more
extensive treatment . Wat is being "read" in this case is the structure
of the passion between Julie and Saint-Peux, which has been acted
out dramatically in the three book that lead up to this concluding
episode.
That "passion" has to be understood as a structural system has
clearly been stated in an earlier recapitulation. Passion is not some
thing which, like the senses, belongs in proper to an entit or to a
subject but , like music, it is a system of relationships that exists
only in the terms of this system: "The source of happiness does not
reside entirely either in the desired object or in the heart that pos
sesses i t, but in the relationship between both . . . " (2:225) . 31 As we
kow from the reading of Narcisand Palion, texts centered on
the specular structure of selfood, this relationship can be stated in
terms of a dichotomy of self and other that engenders a chain of
contrasting polarities. Having moved through such a process of
fgral substitutions, Julie describs it as based on the presumption
of an analog between body and soul , between outside and inside. "I
thought 1 recognized in your face the traces (traits) of a soul which
31 . The statement is made within a context that bring together several of the
novel's main polarities, and heavily engged i n the bad faith of what Russeau calls
"morale de pratique. " It is used by Julie i n an antihedonistic sermon to prove that
Sint-Preux would be ill advised to substitute pleasure for love. Pleasure, as the
satisfaction of needs, is indeed not founded on relationship but on possession. That
her own plea is nevertheless dependent on the seductive voabular of pleasue is
clear, but dos not permit us to discard the statement .
ALLEGORY (JULIE)
211
was necessar to my own. It seemed to me my senses acted only as
the organs of nobler sentiments, and I loved you, not so much for
what I saw in you as for what I thought I felt in myself' (2: 340) . The
chain of substitutions crosss from "visage" (outside) to "are" (in
side) by way of "traits" ("les traits de l'ame") which are said to b
both inside and outside, an ambiguit made plausible by the fact
that , despite their exteriorit, the lines ofa face produce a semiolog
cal as well as a physical connotation, and appear as the inscription,
on the surface of the face, of the soul's meaning. Simultaneously, we
pass from "sens" and "yeux" (outside) to "sentiments" by means of
the synecdoche "orgne" ("mes sens ne seraient que d'organe i des
sentiments . . . ") . "Voir" and "sentir" also accomplish the transfer
ence to the categories of self and other, since now to "see" the other is
to "feel" the self ('Taimai dans vous, moins ce que j'y voyais que ce
que je croyais sentir en moi-meme") . This transfer occurs because
the slfis said to b in need of the other, to b lackng something that
only the other can provide: "les traits de l'ame qu'il fallit i la
mienne. " The void, the hole, is flled, as it were, by the soul of the
other, which is of course also his body. The dynamics of the chias
mus require a valorization achieved by calling the desired feelings
"noble" ; the continuity from sensor to ethical hierarchies is part of
the same metaphorical and analogical system. But the crossings have
upset the authorit of the construction and produced a constant
emphasis on possible delusion: 'e crus voir . . . "; "il me sembla que
. . ; "; 'e croyais sentir . . . "; modalities that throw their shadow on
the parallel verbal construction extending from 'e crus voir . . . " to
"il me sembla que . . . " to 'e croyais sentir . . . " and fnally to
'Taimai en vous. . . . " Under the impact of so many mental resr
vations, the verb "aimer" almost acquires an optative tone and it is
indeed soon bluntly stated that all these substitutions were gounded
in an aberration which now belongs to the past: "Less than two
months ago, I still thought that I was not mistaken . . . " (2: 340) .
Love must now be called blind ("l'aveugle amou") in a way that
deviates considerably from the commonplace associations with the
blindfolded Cupid, all the more so since blindness is stated wthin a
(negative) context of truth and falsehood: "l'aveugle amour, me
disais-je, avait raison . . . " (2:340) . The self-destructive power of the
passions is not due to outside causes but is grounded in unwarranted
assumptions abut the coherence of a world in which the re
semblance of appearances would warrant the afnit of essences.
The passage explicitly discards the notion that the evil consequences
212
ROUSSEAU
of passion might be the result of external obstacles, such as social
prejudices or parental tanny: "I still thought that 1 was not mis
taken; blind love, 1 told myself, was rght; we were made for each
other; 1 would be his if the human order had not upset the order of
nature . . . " (2:340). From a narrative point of vew, the statement
shifts the pattern of rferential authority from a representational,
mimetic mode (which has been prevalent in the unfolding of the
novel up to this point) to a deconstructive diegesis. For up till this
letter, the interest of the action had been primarily based on ele
ments now found to be fallacious: julie and Saint-Preu have been
presented as stock characters in a situation of sentimental tragedy,
persecuted by the social inequities of wealth and class and by the
caprices of a tannical father. The reader's responses are solicited
according to the rules of this plot, thus maintaining the homolog
between enunciation and understanding that characterizes monolog
ical narratives. With the discover, in retrospect, that this symmetr
is an illusion, the entire narrative has to be reconstructed along
different lines. The reading has to check itself at all points, in quest
of clues that puncture the surface of the discourse and reveal the
holes and the traps concealed underneath. Rading now requires a
vgilance that can no longer simpl trust what it hears. Areas of the
text obscured by the succession of predictable events and feelings
become again apparent as a new network of narrative articulations
replaces the frst. What appeared at frst as a sequence of lyrical
moments, separated from each other by the well -rounded closure of
each particular letter, becomes, in the recapitulation, a narrative
chain of successive errors, as misleading for the reader as they were
for the character-not unlike the steady degradation from invention
to historical catastrophe by which, in the Son Diourse and in the
Essay, manknd has been brought down to its present condition.
Like all metaphorical sstems, the frst part of L nouvell
He10fe (Book 1 to 3) consists ofa chain of substitutions and, as in all
deconstructive narratives, the second reading, called forth by the
recapitulation, reveals the weakess of the links by means of which
the polarities were held together. The relationship between julie and
Saint -Preux is told as a substitutive movement in which self and
other constantly exchange their identity, as if the were a single
androgous being whose unit could not be deranged by the inter
nal transfers of attributes: "Come and unite wth your own self'
("Viens te reunir a toi-meme" [2:146]) ,julie tells Saint - Preux, and she
is for him like the omnipresence, the parousia of an element fner
ALLEGORY (JULIE) 213
than air: "I see you, I feel you everhere, I breathe you in the air you
have breathed; you penetrate my entire substance" (2: 147) or, in
another "Platonic" passage,32 "I would imagne you to be of a higher
species, if the devouring fre that penetrates my substance did not
unite me with yours and made me feel that they are the same"
(2: 116) . It is not surprising that their sexual union is linked from the
start with an imager of incest . Not ver long before their frst em
brace, Saint-Preu will say: "I would tremble to touch your chaste
body as ifit were the vilest incest, and your safety is not less inviolate
in my company than it would be in your father's" (2:42) . "I am no
longer in possession of myself," Saint-Preux had said somewhat ear
lier, "my estranged soul is entirely wthin you" (2: 101) , and in re
sponse]ulie can see herself as totally immolated before and replaced
by the other: "Be my entire being, now that I am no longer anything
. . ." ( 2: 103). These self/other substitutions are familiar from Nar
cise and P
l
in and can be transfered from the relation b
tween the lovers to the relationship between author and work and
fnally between author and reader. A sentence like the following
could have come dirctly out of Pglin or out of the tradition that
stands behind i t: "You have lef in me something of the inefable
charm that inhabits you, and I believe that wth your gentle breath
you inspired a new soul within me. Hasten, I beg you, to complete
your work. Take whatever remains of my soul and put yours entirely
in its place" (2: 150) .
I f the substitutions could indeed occu within the totalit of a
single andrognous being, they would still engender other metaphors
b which this unit would be asserted, be it in the form of myhs or
ofimages. A full cosmos is an inexhaustible reseroir of complemen
ta symbols. But the polarties that stem from an illusor plenitude
wll lead to ever-wdening dissonances and generate a ver diferent
type of stor. The frst three books ofuli are a tical version of
such a stor. They include all conceivable confgurations of the origi
nal oppositions, intertwined in patterns that are never allowed to
stabilize, for whenever a substitution has taken place a new unbal
ance, by excess or b default, is revealed and requires new displace
ments.
It would be too lengthy a task to trace in detail the chain of
transformations that make up the narrative segment ending with the
last letter of the thir part , but even on the simplest thematic level it
32. The reference is, of cours, to Sympium, 189, 190.
214
ROUSSEAU
is clearly in evidence. As a spatial , topographical structure, the frst
half of L nouvelle He1ie is a succession of uncoordinated, erratic
movements, a series of fights and retuns put in motion by the ver
frst sentence of the narrative: "il faut vous fuir, Mademoiselle, je Ie
sens bien . . . . " Saint-Preux's incessant comings and goings, from
the Valais, to la Meillerie, to Paris, fnally literally to the end of the
world, are reminiscent of the restless pacing up and dow of Pg
malion in his atelier. The difer frm Plion however in that the
motions are not comparable to the dance of a swarm of insects
around a single light , which, in this case, would bejulie. This would
invest all attributes of being into one ofthe two poles and thus create
a ver diferent , singe-centered system. Even the few examples men
tioned above indicate that the substitutions cannot occur in one
single direction and that they have to travel from Saint-Preux tojulie
as well as in the opposite direction. Saint-Preux's geographical agita
tion has its counterpart in the vacillations and "langueurs" ofjulie's
state of mind; between his "outside" and her "inside," there develops
an interplay of complicated and by no means balanced exchanges.
The temporal pattern is equally unstable'. Sparated from julie
by the breadth of the lake, Saint-Preux writes to her the tpe of ode to
the glories of the moment which is so familiar in the Petrarchan
tradition. A recurrent theme exalting or denouncing the seductions
of the moment runs through the book, reaching at times high points
of lyrical intensit; speakng of his "days of pleasur and of gor
with julie, Saint-Preux can say: "A gentle ecstasy absorbed their
entire duration and condensed it into one point , like the duration of
eternity. Neither past nor future existed for me and I could taste
simultaneously the delights of thousand centuries" (2: 31 7) . The exal
tation of the moment is counterbalanced by the contrasting appreci
ation of duration for its ow sae: "sentiment dies away with time,
but the sensitive soul remains forever" (2: 16) . Duration is indeed the
privleged temporal mode for a system in quest of its ow authorit
and striving for a state in which events ae no longer changes but the
confrmation of an identit no longer threatened by an exterior force.
It is the proper tempral mode for a lengthy, monodic, and even
toned narration towars whichluli seems, at times, to b tending, a
narrative in which the void of signifcation would no longer be ex
perienced as a loss. In an allegor of this knd, duration has to be
valorized as the attraction of what is kow to be least attainable.
The text does not however describe this dialectic of instant and
duration, of sameness and of change, in which the seductions of the
ALLEGORY (JULIE) 215
moment also acquire those of repse, as leading to a stable sythesis.
Duration, the coincidence of an entit wth its ow present , requires
the vocabular of an inwardness detached from anyhing that is
other or elsewhere, containing nothing desirable that is not already
possessed. It evokes a fulfllment no longer associated with desire,
since desire is organized around the moment that separates posses
sion from its opposite. Saint-Preu describs "the time that follows
the time of pleasure" (2: 149) , the transfer of attributes derived from
desire into a new, tranquil state: ". . . I adored you and desired
nothing. I could not even imagine any other bliss . . . What tranquil
ity in all my senses! Wat pure, continuous, universal voluptuous
ness! The source of pleasure was in the soul ; it never lef it, it lasted
forever. What a diference between this peaceful state and the agita
tion of love!" ( 2: 149) . Duration appears as self-enclosed and auton
omous, yet it borrows 'jouissance" and ''volupte'' from a restless
outside world governed by "les fureurs de l'amour. " It remains lined
to this world by sensations and memories ("There is a time for
experience, another for memor" [2: 16]) , and it is by ways of this
metonymic link that the metaphorical illusion of duration is
achieved. The ambiguit fully appears when consciousness, as dura
tion, has to realize that it can come into being only at the expense of
the passion that produced the experience of inwardness in the frst
place. At the end of the paragraph from which the prevous quota
tion is taken, Saint-Preux has to ask the by-no-means rhetorical ques
tion: "Tell me, julie, whether I did not love you before, or whether it
is now that I no longer love you?" The exchange between properties
of stabilit and of change engenders an unhappy consciousness: it
occurs in a state of dejection, "in self-shame and self-humiliation"
(2: 149) , and it leads to Saint-Preux's later statement that " . . . we
return to life in order to return to sufering, and the consciousness of
our existence is for us only a consciousness of pain" (2:326) . This
mood is obviously not compatible with a state of duation and re
pose. Neither can it compromise by substituting memories for pres
ence, or by the aesthetic contemplation of its own soul, made "beau
tifl" by the sacrfce of the passion that created it. L nouvll Heli
would be a ver diferent (and a much shorter) text , more like
Werthr or the Mignon chapter in Wilhlm Meistr or SylVi, if the
narrative had been allowed to stabilize in this way. 33
33. As sugested by Claire: "Vous vous direz (apres avoir fait Ie sacrifce de
votre amour) je sais aimer, avec un plaisir plus durable et plus delicat que vous n'en
216 ROUSSEAU
The reversal of the allegor (peripeteia) occurs, then, as the
deliberate rejection of the system of analogical exchanges that has
structured the narative of the novel's frst half. The reversal is no
longer, as was the case in the Scon Dicours, left implicit in the
declining movement, "la pente inevitable" (2: 353) , of the narrative of
eror. It asserts itself as an outspoken decision that sharply divides
the novel in two segments, a before and an afer, entirely modifes
the circumstances and the setting and allows for a retrospctive
vision of remarkable lucidity. The allegor of unread abilit begins by
making its pre-text highly readable: there is not a singe episode,
practically not a single word, in the more than hundred letters that
come before the turning point that is not clarfed and accounted for
by the redoubled reading that the reversal compels us to undertae.
Nor does the narrative hesitate to draw conclusions from the discov
er of its earlier aberrations. In the place of "love," based on the
resemblances and substitutions of body and soul or self and other,
appears the contractual agreement of mariage, set up as a defense
against the passions and as the basis of social and political order.
This decision acquires its moral dimension from the fact that it
moves against the "natural" logic of the narrative and of its under
standing.
This, to, is not selfevident. For it appars that the frst efect of
the decision is one of clarifcation, permitting a coherent interpreta
tion of the frst half of the novel . The complexty of the passage,
which also mark the transition to the allegorical mode, stems from
the fact that, at the moment when Julie acquires a maximum of
insight, the control ovr the rhetoric of her own discourse is lost, for
us as well as for her. The retrospective clarity gained at midpoint
does not extend to the second part : no equivalent recapitulation is
possible at the end ofuli, for it can be shown that the religious
language of the last chapters is nowhere held up as being free of
deluion, i the way the begn g of Ltter 18, Part III, can b sid to
gouteriez a dire: je possede ce que j'aime. Cr celui-ci s'use a force d'en jouirj mais
I'autre demeure toujours, et vous en jouirez encore, quand meme vous n'aimerez
plus. " The supplementar economy of an aesthetic of sacrifi ce (as when Claire
calculates, in the sme passage that "Ie vertable amour a eet avantage . . . qu'i1
ddmmage de tout ce qu'on lIi scrife, et qu'on jouit en quelque sorte des priva
tions qu'on s'impos par Ie sentiment meme de ce qu'iI en cute et du motif qui nous
y
porte" [2:320, my italics]) is expsed when, instead of to sublimation, we move on
to deconstrction, as was already the case in Plion.
ALLEGORY (JULIE) 217
be. The readabilit of the fst part is obscured by a more radical
indeterminac that projects its shadow backwards and forwards
over the entie text . Deconstructions of fgal texts engender lucid
naratives which produce, in thei tun and as it were within their
own texture, a darkess more redoubtable than the error the dispel .
In this text (Letter 18, Pat III) darkess falls when it becomes
evdent that julie's language at once repeats the notions she has just
denounced as errors. Not only does she continue to use a metaphori
cal diction (which would be, by itself, oflittle consequence, since we
are now fully aware ofits dangers) , but she construes the new world
into which she is movng as an exact repetition of the one she claims
to have lef behind. Ifthis is s, then it can be said that julie is unable
to "read" her own text , unable to recognize how its rhetorical mode
relates to its meaning. 34 The repetition difers from its earlier version,
not in structure, but by a thematic shift : it moves from an erotic to
an ethical and religious vocabular, to the odd stratifcation of
pragmatic, practical rtason with a language of high morality and
desire that we frst encountered in the Second Preface toJuli. Virtue
is refered to in the most practical of terms,35 yet it is also spoken of
in a language of religious awe that had hardly been heard, up to this
point, in the novel and that has led critics to speculate whether or not
Julie is supposed to have undergone a religous conversion. Actually,
there is nothing in the strcture ofJulie's relationship to virtue or to
what she calls G that does not fnd its counterpart in her previous
and now s rigorously de mystifed relationship towards Saint-Preux.
This relationship was based on the metaphor of a subject that
34. B the play of notes which allows him to acqui a distancing perspctive
wth regard to juli, Ruseau may seem to escape from this obfuscation at the
expense of his character. But this pattern is anticipated in julie herself, whose
lucidity with regrd to her past experience is never in question and who i capable of
the same distance toward herself as Rousseau allows himself towards her, yet re
mains entirely unable to avert the repetition of her erors. R.'s statement , in the
Preface, of helplessness bfore the opacit of his ow tet is similar to julie's relapse
into metaphorical moels of interretation at her moments of insight. The manipu
lation of point of view is a for of infnite regess inscrbed within the metaphor of
selfoo.
35. Adulter, for example, is denounced for the most practical of reasons:
father should not be forced to support children the may not have sired, flnily life
is disrpted by the constant necessit to lie and to cheat, the continuit of the
succession may be upset, and one should not brea the prevalent rules too lightly
since it is so hard to los the habit once one has done so.
218
ROUSSEAU
difered from another as plenitude difers from lack, and that was
able to exchange the shortcomings of the one for the excess of the
other because a basic afnit compelled them to enter into a rela
tionship of reciprocity. The text insists at length on the need for such
respnse: "The love I have known can only b born from a reciprocal
afnity and from a harmony of souls. One cannot love if one is not
loved or, at least , one does not love for long" (2:341) . Reciprocity or
resemblance is made manifest by the analog between inside and
outside that allows julie to recognize the afnity btween herself and
Saint-Preux by merely looking at his face. Having discarded this
moel , julie at once has to invent an entity called Gd in order to
repeat what she had condemned. G has to be entirely unlike her
self in his self-sufcienc and omnipresence: "Nothing exists but
through Him that is" (2: 358) and, as such, "no model can exist
among incarnate beings to which [He] could be compared" (2: 358) .
Yet he is at once anthropomorhized by givng him the ver attri
butes that made the substttive exchanges with Sint-Pux psible:
a language, a voice, and a face. Lie Saint-Peu, he exists as a
combination of traces ("traits") which make it possible to "read" his
substance: " . . . all his featues [traits] linked to his infnite essence
are always represented to reason, and help reason to restore what
eror and imposture had distorted" (2: 358) . The confguraJion of
these "traits" make up a face, the "inward efg (2: 358) which is
also able to spea with "the secret voice that never stoppd its mur
mur in the depth of my heart . . . . " But this voice can have no
greater authority than the voice of Saint-Preu, since its comprehen
sion depends on the same rhetorical code that proved fallacious in
the fst instance. Consequently, it will be difcult to tell apart the
discourses addressed to Saint-Preux from those addressed to G or
to virtue. Both are based on the same "eternal simulacrum" (2:223) ,
"divine model" (2:224, 2: 358) , "image that we carr wthin our
selves" (2: 358) , and that we are able to prceive and to emulate: "as
soon as passion allows us to see i t, we want to resemble it . . . "
(2:224) . julie and G bcome the two-sided exchange ofa dialogue
in which the words carr shared substances that can be ofered and
receiv
ed: julie's prayer, for example , far from being a radical loss of
selfood before an unintelligible otherness, addesses a kind of over
self that does not difer from her in kind. Attributes circulate freely
within the transparenc of a representational model of expressive
voices and faces: "I see him, I feel him; his helping hand . . . restores
ALLEGORY (JULIE) 219
me to myself, in spite of myself' ; "I want , 1 told him, the good that
you want, and of which you are the sole source"; "Render all my
actions akn to my constant will which is also your will . . . . " Wat
julie wishes to receive from Gd are the same attributes of selfood
and of will ( the prayer repeats 'Je veu" si times, like an incanta
tion) that she requested from Saint-Preux, "the soul which was
necessar to my own" in order to achieve individuation. And she can
again identit this received selfood by recognizing the signs through
which the divinit manifests itself: a face, a voice, or most efectively
of all , certain emotions36 that postulate a continuity between these
signs and their signifcation, just as her own sense of selfood could
be read of from Saint-Preu's countenance. The concatenation be
tween self, feeling, sign ( trait) , and outer appearance (visage) is a
constant network in julie's relationship to God, as it was in her
relationship to Saint-Preu. It is therefore not surprising that , in an
apostrophe frst addressed to Saint-Preux, julie's exclamation, "Ah, 1
have learned too well what it costs to lose you and will not forsake
you a second time! " (2:355) , could be directed just as well to her lover
as to the actual grammatical antecedent in the sentence, divne vir
tue. Neither is it surprising that virtue will later be identifed, by
Wolmar, as being a passion among others, with a structure similar
to that of love (2:493) .
36. It wll not do to interpret the love for Saint-Preu "platonicall as a
prefguration of a transcendental love temporarily directed towards an imperfect
being: the exposure of the rlationship is too radical and the diference btween the
frt and the second part of the novel too wde to allow for such a reading. Brnar
Guyon may be rght in pointing out that a degee of similarit prevails in the
analogical strctures giving rise to metaphors ofinwardness, self will, joy, etc. , that
keep occuring in the scond half of the novel. But this ver similarit is then recog
nized as a patter of eror that remained hidden in the frst three books; instead of
unifing the totalit of the text, it undoes whatever illusion of unit the frst half
tended to conve. Nor is the opposite reading more convncing: Julie's passionate
addresss to G cannot b interpreted as a simple confusion of the divine wth the
erotic, a delusion akin to that of a "quixotic" character lie Emma Bovar. The rigor
ofJulie's insight into the abrations of "romantic" love fnds no remote equivalence
in Flaubert's heroine, neither does the ensuing control ofFlaubert over his fction fnd
an equivalence in Rousseau's confusion wth regrd to his. The problem is not that
Julie remains mystifed, but that a totally enlightened language, regrless of
whether it conceives of i tself as a consciousness or not , is unable to control the
recurrence, in its reader as well as in itself, of the erors it exposes. Julie, the best
conceivable crit ical reader, is apparently unable to read her own critical text criti
cally.
220
ROUSSEAU
Thus the text of the pivotal letter that concludes the frst part of
Juli, s clarifng as a recapitulation, bdes little goo for the stabil
ity of what it proleptically announces. It will be followed by the
lengthy description of the political order in the communit of Cla
rens, of which it is difcult to decide whether it is an exemplar
model for a state or an ambivalent family romance. Wen the lan
guage of selfhood returns, as in the fnal letters again centered on the
relationship between Julie and Saint-Preu, it is in terms that are not
political and not even primarily ethical , but religious. Our reading
tries to account for the emergence of the ethical valorization but
remains unable to answer the question raised by the interpreters of
Juli as to the relationship between the political aspects of Clarens
and the religious considerations in Julie's last letters. The question
could be dealt with by an extended reading ofuli in its entiret, but
since we dispose, in Russeau's work, of at least two more systematic
treatises involvng religous and political theor ( the Pfsion dfoi
from Emil and the Sil Contat) it may be legitimate to follow
his own hint and take him at his word in his assertion that the
Pfesion d foi and the concluding letters of Juli "are in close
enough accord to make it possible to explain the one by means of the
other. "37
37. Lt d l mnt in Ouve cmpl, 3:69.
10 Alegor of
Reading
(Pfeion t foi
WHEN THE SAVOYARD PRI EST HAS COMPLETED THE
theoretical part of his discourse, his listener is impressed enough b
the feror of his tone to compare him to Orpheus, but less inclined to
praise the originalit of a doctrine conveyed with such inspired con
vction: "The views you have just exressed, I sid, seem more u
usual to me because of what you admit you do not kow than
because of what you say you believe. I see in them something ver
close to theism or natural religion, which Christians profess to
equate with atheism or irreligion, though it is actually the exact
opposite" Cp. 6061) . It is indeed well kow that even Voltaire, de
spite many reserations, expressed ageement with the doctrine of
the Pofesion d foi and that William Blake could, wth some justice,
lin Voltaie and Russeau together as promoters of the natural
religion he despised. On a frst level of understanding, the Pofsion
dfoi is unquestionably a straightforard theistic document, 2 basing
religious conviction on the manifestation ofinnate and natural moral
1 . All page refernces are to J. J. Rusau, Ouv complt, ed. Brrd
Gagnebin and Marcel Rymond ( Pari: Galimard [Bibliotheque de l PIIiade],
199) , vol . 4. In this edition, the text of Emil (of which the Psion dfi i a par)
has ben established b Charles Wirz and the commentar by Pere Burgelin. The
English translations from the Pfsin d fi are thos of Ll Bir in Th
Etl Rusu (New York: Rndom House, 1974) but page references are to the
French Ouve compltes.
2. For Ruseau's defnition of theism, se the Sl Cont: the religon of
man, as opposed to the religon of the citizen, i ". . . without temples, altars or
rites, limited to the purely inward cul t ofthe Supreme Dit and to t he eteral dutie
of moralit. It is the pure and simple religon of the Gspl , tre theim and what
can b called divne natrl right"-J. J. Ruseau, Ouv compl, ed. Brrd
Gagebin and Marcel Rymond ( Paris: Gallimard [Bibliotheque de la PIiade],
19), 3: 4. The text and commentar of the Sil Conta, in this edition, ha
been established by Rbrt Drathe.
221
222
ROUSSEAU
feelings: "Let us limit ourselves t o the frst feelings we fnd within us,
since inquir always brings us back to them, when it has not led us
astray. Conscience! Conscience! Divne instinct , immortal and celes
tial voice! You are the sue guide of a being who is igorant and
confned, but intelligent and free. You are the infallible judge of good
and evil and it is through you that man resembles Gd . . . " (600-
-01) . As for understanding what this voice is telling and where this
guide is leading us, there seems to be no serious difcult: "It is not
enough to know that this guide exsts: we must also be able to
recognize and follow it. lfit speak to all hearts, why are there so few
who understand it? It is because it spea to us in the language of
nature . . . " (601) . This is presumably the same language in which
the Book of Natur has been witten: "I was never able to belee that
Gd had ordered me, under penalt of damnation, to be so learned. 1
therefore closed all books. There is only one that is open to all eyes:
the book of nate" (62) . Hence the sublime natural setting chosen
for the vcar's discourse: "Natue seemed to display all its mag
nifcence to our eyes, as if to ofer it as the text of our dialogue" (565) .
The t of nature has its equivalence in the "inner image" [tf g
intrieure] whose voice speaks clearly enough if only we are wlling to
silence the distracting noises of worldliness: "Let us look into our
selves my young friend! . . . There is within our souls an innate
principle of justice and virtue by which, in spite of our maxms, we
judge our acts and those of others as good or bad, and it is this
principle that 1 call conscience" (598) . To the vain teaching of phil os
ophy, the vcar opposes the innate wisdom of "inner ligt" (569) .
The afrmation of belief in a natual religion, founded in the
transcendental valorization of such concepts as inwardness, innate
ness [i
nit] , voice, natal language, conscience, consciousnes, self
hood, etc. , represents, of couse, a considerable deviation from the
epistemological and rhetorical critiques that are to be found
elsewhere throughout the work of Rousseau. No such considerations
play a part in the historical and political arguments of anthropologi
cal texts such as the Son Diourse; when Gd is mentioned in the
Diourse it is primarily to demonstrate that there is no such thing as
a natural language. 3 The confict emerges perhaps most dramatically
3. "Watever the orgns [of lnguage] my have ben, one can s how prly
Natue has prepared men for life in soiety from the l ittle care it tok to brng them
together by natral means or to facilitate thei use of spech" (Oeuve comple,
3: 151) . The same is said abut the development of technolog: men had to lear
ALLEGORY OF READI NG (PROFESSION DE FOI) 223
when one jutaposes the vcar's natural piety wth the pre
Nietzschean denunciations of Christianit, as a political force, in the
chapter on civil religion (Book I, chapter 8) of the Sial Contrat:
"It was in these circumstances [paganism havng become the univer
sal religion of the Rman empire] that Jesus came to establish a
spiritual kingdom on earth. By separating the theological system from
the political sstem, this destroyed the unit of the state, and caused
the internal divisions that have never ceased to agtate Christian
nations . . . . Wat the pagans feared fnally happened. Everhing
then became diferent ; the humble Christians changed their tone,
and the supposedly otherorldly kingdom, under a visible ruler,
soon became the most violent despotism in this world. "4 "The [reli
gion of the priests . . . which includes Rman catholicism as a major
instance] is [considered politically] so obviously bad that amusing
myself by demonstrating its drawbacks would be a waste of time.
Anything that break social unity is worthless. All institutions that
place man in contradiction wth himself are worthless. "s Jean-Robert
Tronchi n, the Procureur general of the Cit of Gneva, whose report
led the Petit Conseil to order the burning of Emil and the Scil
Contrat (as "tending to destroy the Christian religion and all gov
ernments") was certainly justifed in referring to these texts in his
cogent attack on Russeau published under the title Lttres d la
Campag.
The gap between the theophany of the Prfsion d foi and the
political writings, especially the Scil Contat Uuli being an inter
mediate case, too complex for summar discussion) , is too obvious to
have escaped notice and remains one of the main cruxes in Russeau
interpretation. The critical rigor of the political texts contrasted with
the piet of religious sentiment in the Profssion d fi always again
forces commentators, depending on their temperament and convic
tions, to daemonize the former or to condescend to the latter. Thus
Pierre Burgelin, in his introduction to the fourth volume ofthe Com
plte Work, says ofthe discrepancies between statement and tone in
the Prfesion that "these disparities stem from the distance between
intellectual analysis and the impulses of the heart" (cxlii)-lways
and again the Schillerian dichotomy between Russeau's Denkkrf
various techniques from the Gs "since it is impossible to conjecture how they could
have taught them to themselves" (Oeuve complt, 3: 145) .
4. Oeuve cmplt, 3:462.
5. Ibid. , 3:464.
ROUSSEAU
and his Empfnlihkit. The usual temptation arises t o account for
the contradictions in various empirical and thematic ways. One can
arge that the movement of Rousseau's thought tends towards reli
gous conversion, a hypothesis that receives some support from bio
gaphical data as well as from the fact that the Nouvll He11 can b
read so as to mae conversion the central statement of the book. Or
one can simply accuse Russeau ofinconsistenc and dismiss the one
or the other half of his schizophrenic speculations. More prouc
tively, one can consider the religious aspect of his work as an ideolog
ical superstructure resulting from the repression or sublimation of
pschological or political contradictions, in a movement that runs
parallel to Althusser's diagnosis of the literar sublimation of Rous
seau's political confusions. Before following any of these sugestions,
one should begin by establishing if, on the basis of the exsting texts,
the discrepancies indeed occur in such philosophically uninteresting
way. In the main tets, Juli, Emil, and the Sl Contra, religous
and political elements are closely intertwined, but their relationship is
far from being either peaceful or simply comprehensible.
On the somewhat more specifc issue of the source of legal and
moral authorit, the discrepanc between the Pfesin d foi and
the Si Contrat is not less striking. In political decisions having to
do with conflicts between the general will and individual interests,
the "inner voice" of conscience, which the Vicar held up as the source
of all trth, is of no avail : "Wen the general will has to be consulted
on a particular act . . . what wll [common man] do to shelter
himself from error? Will he listen to the inner voice? But this voice is
said to be merely the result of judgments and feelings within the
sphere of an existing societ and according to the laws of this soiety.
I
t can therefore not be used to establish these laws . Moreover, if it is
to b audible, none of the passions that speak louder than con
science, that cover up its timid voice and allow philosophers to claim
that this voice does not exist , should have arisen in the heart of
man. "6 This sme "shyess" of conscience (although it is the voice of
G) is stressed in the Pfsin d fi (and in the letters to Sophie
d' Houdetot which are closely connected to the Pfsion) : "Con
sciousness is shy, it loves retreat and peace . . . " and, in the long run,
i t can even fall silent altogether: "I t no longer speaks t o us or answers
us, and when we have despised it so long, it becomes as difcult for
6. Sil Cnta, frst version, Ouve complt, 3:27.
ALLEGOR Y OF READI NG (PROFESSION DE FOl)
us to call it back as it was to banish it" (601) . 1 At other times,
however, the same voice can be so loud that only the deaffail to hear
i t: "[The materialists] are deaf to the inner voice that cries out to
them in a tone that is difcult to ignore: A machine does not thin
. . . " (585) . A geat deal of ambivalence, historical , metaphorical ,
and logcal , thus blus a point that , according to all students of
Russeau, is crucially important in the interpretation of his thought.
Indeed, the entire question of the relationship btween the gen
eral will and particular volition, beten public and private morality
Lter to Fanu;r) , between public and private well-being (Du
bonhur publi) , between the theologcal and the political order,
hinges upon the uncanny timidity of the divine voice. Is G timid
because he is so exquisitely sensitive or because he is not himself
quite certain of what he has to say? There certainly is nothing timid
about the laws which are supposed to be dictated by this voice and
which confer, for exmple, the right of life and death over indivdu
als (Sil Contat, II, chapter 5) . 8 It is difcult to see how Pierre
Burgelin, as one instance among others, can slide so easily over the
difculties. Having to account for a similar discrepanc between the
frst and the second version of the Sl Contrat on the source of
legal authority, he writes: "These two texts seem to contradict each
other. But this is not the case: universal justice orignates in reason
enlightened by G [y conscience] , and it only becomes applicable
in the law. Thus the law changes from being a celestial voice to the
condition of its applicability; the celestial voice is transferred to
moral consciousness and fnally to God" ( 1562) . Burgelin moves
from G to "voice," from divine voice to human conscience, from
human conscience to practical moralit, from moralit to political
law, in a sequence of mediations threatened at all points b number
less aberrations. For a mind as distrustful as Russeau's, little in
clined to have faith in any voice including his own, it seems unlikely
that such a chain of displacements could be mastered without
further complications.
The mere jutaposition of explicit statements or the recourse to
extra-textual causal exlanations does nothing to resolve the difcul
ties. Since the question focuses precisely on the possible understand
ing or misunderstanding of voices that
r
state other voices, meanings
7. Amost the same formulation appears in Let ter 6 of the Ltmral (l
Sphie d'Houdetot ), Oeuve cmpl, 4: 1112.
8. Oeuv complt, 3:376-77.
226
ROUSSEAU
that reread or rewrite other meanings, the reading of the Pfsin
d fi should be less monological than has been the case, with ver
few exceptions, in the interretation of Russeau's religious and
political writings. It is obvious, for example, that none of the pas
sges so frequently quoted from the Pfsin as evidence for Rus
seau's theistic convctions are spoken by Rousseau himself, but by a
fctional character, whose "voice" does not necessarily coincide wth
the author' s; the same is true, of course, in Julie's letters. In the case
of a so-called work of fction, the obseration is almost too self
evdent to be necessar; no one will ; without further question, simply
equate Poust with Marcel or Flaubert with Emma Bva. But in a
discursive text lie the Pfsin d fi, the use of a fctional spokes
man, if it is noticed at all, is explained empirically as an alibi to
shelter the witer from reprisals for his subversive opinions, a real
enough problem in the case of Emil. Yet the presence of a fctional
narrator is also a rhetorical necessit in any discourse that puts the
truth or falsehood of its own statement in question. More still than
epics or novels, discursive texts are necessarily dialogical-which
implies, among other things, that they cannot be qutd without frst
havng been rea. The unwaranted separation between the way of
reading and interreting "literar as opposed to "philosophical" or
discursiv texts-a separation due in large measure to ideologies
derived from the misuse of aesthetic categories-eprives the read
ing of philosophical texts of elementar refnements that are taken
for granted in literar interretation. Paradoxcally enoug, this
seems to happen even more clearly in the case of rhetorically self
conscious writers like Plato, Russeau, or Nietzsche than in that of
more formally technical philosophers.
The quotations that support the reading of the Pfsin d foi
as a defense of natural religion frame an extended argument, one of
the most sustained philosophical developments in the entire Rous
seau corpus (567-60) . The argment is in part polemical and
primarly directed, as is well known, agaist Helvtius's D ['eprt , the
article "Evidence" in the Enlped, and, beond that, the or
thodox of a materialistic interretation of nature associated with
the work of Buion, d'Holbach, Maupertuis, la Mettrie, and certain
aspects of Diderot. 9 The place of the Pfsin d fi in the intellec-
9. The soures of the Pfesion d foi have ben extensively studie, frst by
Pere-Maurce Masson in L "prfesin dfoi du vair &vyar" d]ean-acqu
ALL EGORV OF READI NG (PROFESSION DE FOn 227
tual histor of the eigtenth centur, as a blated Cartesian teeo that
resists both the Spinozistic and experiential trends of the times, has
been well documented and concerns us here only on the frst level of
understanding. Indispensable as the are, such historical considera
tions have been an obstacle to the reading of the Pfsin, if only
because the promote a tendenc to oppose the sentimentalism of
the pious passges to the rigor of enlightened rationalism and thus
impose upon the text a binar system of valorization that is alien to
its structure.
The argument ofthe Pfsin d fai begins ( 567) in Montaigne
rather than in Descartes, in a condition of radical doubt that is more
empirical than epistemological . As is consistently the case in Rus
seau, the reduction to a condition of mere self-presence, be it as an
individual consciousness or, as in the Esai sur l'arigne d lang,
as a political societ, does not result in a constitutive cagta. It is a
moment of genuine and intolerable confusion that allows for no
statement other than its ow intolerabilit. Consequently, it is more
liely than any other moment to lead astray: tortured by doubt , "[the
human mind] would rather be mistaken than believe nothing" (56) .
But since the original confusion is itself caused by error, by the
inaccessibilit of truth ("I love truth, I search for it but fail to recog
nize it . . . " [567]) , the addition of more error to an existing state of
aberration is not likely to improve things. The invocation of Des
cartes's name ( 567) has from the outset placed upon the argument
ar epistemological burden that makes it impossible to valorize such
terms as "eror" or "illusion" in a positive way. Russeau can then
reiterate the classical gesture of a tabula raa and reject all exsting
wsdom as the product of mere conceit . But whereas this gesture
should traditionally be a preliminar to the counterassertive integrity
of self-reflection, it fails to lead, in this case, to any such assurance.
The only clai m made for the "inner light" that the mind is able to
throw upon its powers is a dubious, unfounded hope for a lesser evi l ,
entirely unable to resolve the condition of uncertaint that engen-
Rusu, edition critique (fribourg and Paris, 1914) . Masson's fnding are dscussed
and completed in Burgelin's introouction and notes to the Ouve comple, with
special emphasis on Helvetius and the article "Evidence" f'om the Erlpe, now
generally blievd to have been witten by franois Quesnay, the best kown of the
"physiocrats" (see 3: 1129, n. l ) .
10. for example in Henri Guhier, r mtatins mtaphysiu d Jean
Jaqu Rusu ( Paris, 1970) , chapter 2.
228
ROUSSEAU
dered the mental activit in the frst place: ". . I must follow the
inner light , it wll mislead me less than [the philosophers] do, or at
least my error will be my own, and I shall be less pererted if! follow
my illusions than if I believe their lies" ( 569) . The shift to ethical
valorization ("me depraver") , sugesting that the distinction be
tween "illusion" and "lie" is primaril a question of Rousseau's good
faith as opposed to the false pride of the philophs, is at the ver
least premature since, at this point , the question at issue is one of
truth and falsehoo and not of god and evil. Rousseau never
claimed that good faith sufces to give authorit of trth to a state
ment or a knowledge.
Neither dos he claim it here. Still guided by the same valoriza
tion that privileges inside conviction over outside opinion, a polarit
that has been introduced by the conventional rejection of all received
knowledge as coming from "outside, " we gide without discontinuit
from sigt to sound (from light to voice) and are told to follow
"l'assentiment interieur" ( 569) in accepting or rejecting the results of
our attempts at understanding. Does this "inner assent" then acquire
the paradigmatic qualit of a Cartesian cogto as the foundation of
judgment? This is hardly the case, for the description of its workings
indicates that the "inner assent" operates only with regrd to "ideas"
that hav themselves been identifed (Le. , understood) by means of
an act of judgment that has nothing to do with an immanent assent :
"Then, turning over in my mind the various opinions that had suc
cessively swayed me since my birth, I saw that . . . my inner assent
was given to them or wthheld from them in varng degrees. Having
made this obseration, I compared all those diferent ideas without
prejudice and found that the frst and most common was also the
simplest and most sensible . . . " (569) . To compare is, for Russeau,
the distinctive qualit of judgent , thus making it clear that the
inner assent is itself dependent on a prior act of judgment which it
does not control . The strcture of the argment is in fact more
deceptive, for the "frst and most common idea" is identifed as being
precisely the theistic claim for the immanent authorit of conscience,
an article of faith here stated allusively by reference to the name of
Samuel Clarke. The only thing to which the "inner assent" assents is
itself; it sets up a tautological structure devoid of the deductive
power inherent in a Cartesian cogto. It is true that we are being
advsed, in the next paragraph (570) , to decide upon the truth of
further units of knowledge deductively, by ascertaining their "neces-
ALLEGORY OF READI NG (PROFESSION DE FOI) ZZ9
sar link" wth the original evidence of the inner assent. But this
appaent deduction is an ilusion, since the necessit of the link can
only be verifed by means of the same rle of evidence that estab
lished the validity ofthe orignal prnciple and thus infnitely repeats
itself wthout modifcation: assentio ergo asentio, etc. -unless one
makes the principle of verifcation into an independent act ofjudg
ment (as was already the case in postulating the possibility of com
paring ideas) , but then the evdence of immanence is no longer the
"frst . " Rgrdless of how one construes the passage, it indicates that
the principle of immanence is in fact being suprseded by an act of
judgment which does not necessaily claim to possess the constitutive
or generative power of a cogto. In this respect , the Pofsin d fi
may well b a pre-Knt ian rather than a neo-Cartesian text. For,
despite the appant confusion of its point of departure, it is not in
fact confused by its own inconsistencies. It immediately draws the
correct inferences from difculties that could well have paralyzed
the argment from the outset. The problem now becomes, not how
to construe an interpretation of exstence by means ofa rule of inner
assent, but to account , by a critical act of judgent, for the ocur
rence of such an assent and to establish its epistemological status.
Russeau acknowledges at once the indeterminac of his own self
reflection by movng into a crtal vocabular: "But who am I? What
right do I have to judge things and what determines my judgments?
If the are forced upon me b the impressions I receive, i t is futile for
me to expend my energ in such inquir, because the wll not occur,
or will occur on their own, wthout any efort on my part to control
them. First , therefore, I must examine myself to come to know the
instrument I intend to use and learn the extent to which I can rely on
it" (570) . The main informing concept ofthis text is that of jugnt,
not inner light or inner assent. The argument of the Pfesin d foi
seres to reveal the structure of judgment in Russeau and to estab
lish its relationship to other ke concepts such as wll, freedom,
reason, etc.
The structure of judgment , in this text , is established by opposi
tion to that of sensation or perception. Moreover, what wl be said
about judgment wll apply, albit wth a diferent and independently
interesting thematic content , to three concepts shown to be correla
tives of judgment: will , reason or intelligence, and freedom. No in
herent priorit or genetic link exists between the four terms (udg
ment , will, reason, and freedom) and it seems that a decision to start
230 ROUSSEAU
out the argment i n terms of judgent i s to a large extent arbitrar.
They form a coherent conceptual chain in which each term can be
derived from the other at wll : "If I am asked what cause determines
my will, I ask in tum what cause determines my jugt: for it is
clear that those two causes are one. Ifwe realize that man is active in
his judgment and that his intlligence is only the power of comparing
and judging, we see that hisfedm is only a similar pwer, or one
derived from i t. He chooses good as he chooses truth; ifhis judgment
is false, his choice is wrong. What , then, is the cause that determines
his will ? It is hisjugt. And what is the cause that determines his
judgment? It is his facult of intelligene, his power to judge . . ."
(58, my italics) .
Rousseau's theor of judgent restates, though in a less oblique
and bewildering manner, the critical theor of metaphor that under
lies the argment of the Scon Dicourse. Judgent is described as
the deconstruction of sensation, a model that divides the world into a
binar sstem of oppositions organized along an inside/outside as
and then proceeds to exchange the properties on both sides of this
axis on the basis of analogies and potential identities. The systems
conceived by the contemporaries against whom Rousseau
polemicizes are all structued according to this fundamentally
metaphorical pattern: by means of empirical considerations on the
nature of perception, they oppose body to soul , sensation to judg
ment , nature to mind (or art) , rs ee to re cogto , outside to
inside, death to life, and then reconcile the antinomies wth varng
degrees of dialectical rigor. For example, in the hylozoistic vision ofa
biologically alive matter, even death and life can be reconciled: to see
the natural world as a live animal is to push to its limi t , as Diderot
will ironically sugest in Pee sur I'interrtatn d l nature,
1 1
the metaphorical model of an inside/outside correspondence. ,Rous
seau, in the discussion of judgment , categoricall rejects the unwar
ranted totaliztion of metaphorized syecdoches: "Yet this visible
universe consists of matter, scattered and dead matter, which as a
whl has none of the cohesion, orgnization or the common feeling
of the par of a living body, for it is certain that we, who are par,
have no feeling of ourselves in the whl" (575, my italics) . Sensation
unadulterated by judgment is in fact inconceivable, but if it is posited
1 1 . Denis Diderot , Ouv cmplt, chronological edition wth introduction
by Rger Lewinter (Paris, 1970) , 2: 767-71 .
ALLEGORY OF READI NG (PROFESSION DE FOI) 231
hyothetically (as the state of nature is posited in the Scond Di
course) , it is totally incapable of setting up any relationship beteen
entities or of setting up entities in the frst place. It would not be
adequate to call the rhetorical structure of such a world of pure
sensation metonymic rather than metaphorical , since such a strc
ture could not even conceive of contiguity, let alone of resemblance.
Rousseau writes as if he admitted the existence of an inside/outside
polarit ("I therefore clearly understand that a sensation, which is
inside me, and its cause or object , which is outside me, are not the
same thing" [571] ) , but this "hors de moi" is then so entirely devoid
of any semblance of coherence or signifcation as to be nothing at all ,
for us or in itself. To a pure sensation, it would appear entirely
chaotic, contingent and unpredictable, much more so even than the
picture of the orgnic world gone awr that Diderot conjures up in
the Lttre sur l aveug. 1 2 A universe of pure senstion would be
unable to conceive of ratio or of number: "Such comparative ideas as
'larger' or 'smaller' and the numerical ideas of 'one,' 'two,' etc. , are
certainly not sensations, even though they are always accompanied
by sensations when my mind produces them" (572) . Pure "outside
ness" is the only discontinuit that articulates even the smoothest
appearance of identity, since it cannot be said that the sensation of a
given entity in X is identical with the sensation of the same entity in
Y: "When the sensations are diferent, the sentient being distin
guishes them by their diferences; when the are alike, it distin
guishes them because it feels that the one is outi the other [hors
des autres]" (572, my italics) . Spatial models-nd the same would
have to apply to temporal models-are metaphorical conceptualiza
tions of dierntal structures, which is why the engender such
redoubtably efective and misleading powers of unifcation and
categorization. Rousseau, however, introduces outside ness (wth its
implicit train of spatial and temporal corelatives) as a prnciple of
diferentiation that could never legtimately be made to fnction in
the opposite direction: the "outside" in the sentence just quoted is not
the outside ofa corresponding inside. In the mode of pure senstion,
everhing is "outside" everhing else; there is nothing but outside
diferences and no integration is possible.
This version of diferentiation is similar to the distinction made
in the Scond Dicours, also on the basis of diference and re-
12. Iid. , 2: 197-98.
232
ROUSSEAU
semblance, between denomination and conceptualization. The act
of judgment coincides with the abilit to postulate relationships, the
possibilit of elaborating sstems based on the correlation of difer
ential with integative moments. This activit is called "to compare"
or, more explicitly, "to place one [object] over another to pronounce
on their diference or on their similarit" (571 ) . The process is a
manipulation, a displacement that upsets the "truth" of things as
the are, for sensation, unlike judgment , is truthful to the extent that
it leaves things in their proper places and does not even conceive of
mang what is distinct identical : "By sensation, objects are pre
sented to me as separate, isolated, as the are in nature . . . " (571 )
and therefore "I kow only that truth is in things, not in the mind
that judges them, and that the less ofmyselfI put into my judgments
of them, the surer I am to approach truth . . . " (573) . Since judg
ment is also associated wth what, in this same paragaph, is called
"the honor of thinkng," it follows that thought and truth are not
necessarily coextensive notions.
Judgment, also called "attention," "meditation," "reflection," or
"thought" [ene] , and always described by verbs of motion such as
to move, to transport , to fold [rplier] , neither reveals things for what
the are nor leaves them undisturbed. It moves them around, thus
mimickng the etymolog of the ver term metaphor, of the Aristote
lian epipfra: "in comparing [objects] , I move and transpose them,
so to speak, I place one oer another . . . " (571) . Judgment does this
in order to create sstems of relationship that are not substantial but
merely structural ; from a formal point of vew, these systems are by
no means arbitrar but since the are devoid of ontological authority,
they are not controlled by considerations of trth and falsehood.
Therefore, they are capable of errors which they make possible by
their ver exstence: "Why then am I mistaken about the relation
between these to stick, especially when they are not parallel? . .
Wy is the image, which is the sensation, unlike its model , which is
the object? Because I am active when I judge, and the operation of
comparison is fault; in judging relations, my understanding mingles
its errors wth the truth of sensations, which show only objects"
(572) . The falsehood dos not have a contingent cause that could be
corrected by trial and error, by experimental or methodological
refnements, since the act of thought is itself, by its ver manifesta
tion, a falsifcation. 13 The pattern of this falsifcation is the same as
13. The statement recurs frequently in Russeau. Se, for example, in Emil:
ALLEGORY OF READI NG (PROFESSION DE FOI)
that of the concept in the Scn Diourse: it uses strctual re
semblances in order to conceal the diferences that permit the ver
articulations of structure.
The consistenc, or repetitiveness, of Russeau's thought on this
point reduces the polemics of the Pfesion d fi against the
philosoph to a secondar function. The disagreement with Hel
vetius, la Mettrie, and to a lesser extent, Diderot, repeats the earlier
disagreement with Condillac and the tradition of Locke on the origin
of langage. 1 4 It is a central insight of Rousseau that never varies
throughout the work and that fnds its fst systematic expression in
the section on language in the Sond Dicourse and in the Esay on
th Origns of Lngge. The same rhetorical vgilance determines
the disjunction between judgment and sensation in the Pfsion d
foi and that between judgment and sensation in the earlier texts. And
the equation of judgment with langage is hardly less clear in the
later theological than in the earlier political text : it is asserted in a
single sentence, when it is said of judgment that "In my opinion, the
distinctive facult of an active or intelligent being is the abilit to give
meaning to the word 'to be' " (571 ) .
The ambivalence of judgment stands fully reealed in this sen
tence. The stability of the natural world is by itself devoid of meaning
and cannot become a source of knowledge. Being is for us only "the
wrd 'to be' '' , and the copula has no transcendental referent b
natural or divine right . This negative insight , achieved in the difer
entiation between judgment and sensation, is itself an act of judg
ment and this act is verbal . Terms such as "to move" or "to fold" are
replaced by such verbs as "to pronounce" or "to give meaning," thus
implying that the verb "to be," as matrix of all referential langage,
has no proper meaning by itself. The scene of judgment is that of a
verbal pronouncement and that of an oracular verdict. But after
having thus undone any possible association between relation and
necessity, the same judgment then proceeds to do, in its own name,
what it had undone in the name of sensation, and to set up struc
tures, such as concepts, which lay claim to meaning in the same way
that sensation could lay claim to the existence of matter and reality:
" . . . the more men kow, the mor the err . . ' . " (3:4) . In contrast to this,
Helvtius, in D l 'eprt, wil l sy: ". . . ou false judgents are the result of acciden
tal causes which do not imply, in the min, a facult of jugent that would b
distinct from a facult ofpivng [facuIte de str]" (quoted by Burgelin, 3: 1523) .
14. Se Chapter 7, p. 152.
ROUSSEAU
"Thus not only do I exist , but other entities exist also, namely, the
objects of my sensations . . . [which] I call matter . . . " (571) . Like
language, judgment engenders the same possibilit of reference that
it also excises. Its error can therefore not be localized or identifed in
any way; one could not , for example, say that the error stems from
language, as if language were an entit that existed independently of
judgment or judgment a faculty that could exercise its activit in a
nonlinguistic mode. To the extent that judgment is a structure of
relationships capable of error, it is also language. As such, it is bound
to consist of the ver fgural structures that can only be put in ques
tion by means of the language that produces them. What is then
called "language" clearly has to extend well beyond what is empir
cally understood as articulated verbal utterance and subsumes, for
instance, what is traditionally referred to as perception.
This becomes apparent in Rousseau's treatment of the Lockean
problem of Molyneux and the senu communum, a recurrent ques
tion in eighteenth-centu philosophy of perception. 1 5 The Pofesion
d foi distinguishes between sensation and perception as Russeau
qistinguishes elsewhere beteen verbal and nonverbal drives
("passions" and "besoins") : "If we were purely passive in the use of
our senses, there would be no communication among them; it would
b impossible for us to kow that the object we touch and the object
we see are the same. Either we would never be aware of anything
outside ourselves, or there would be for us fve perceptible substances
whose identit we would have no means of perceivng" (573) . The
unit of perception is an act of judgment which, as such, denies that
the totaliztion of perception could be rooted in an exchange of
properties held in common by mind and matter. While denying
therefore that perception could ever be an access to true knowledge,
the passage ackowledges that it is structured like a metaphor and
thus must be considered as an act of judgment or a language. The
term "language" thus includes that of perception or sensation, imply
ing that understanding can no longer be modeled on or derived from
the experience of the senses; in this respect , the Pfesion d foi
occupies a pivotal position in the complex histor of the relationships
between empirical psycholog and theories of language in the
ei
ghteenth centur. We can conclude that the vcar describes judg-
15. John Locke, An E y Conering Humn Undrtandin
g
, 00. Alexander
Campbell Fraser ( New York, 1959) , 1 : 186-87; Book 2, chapter 9, 8.
ALLEGORY OF READI NG ( PROFESSION DE FOl) 235
ment as the power to set up potentially aberrant referential systems
that deconstruct the referentiality of their own elaboration. This
description warrants the equation of judgment wth fgural lan
guage, extensively conceived.
Diferent versions of the same aporia organize the description of
the related concepts of wll , mind, and freedom. In the discussion of
the will (576-78) , the polemical argument against Toland and la
Mettrie denying the immanence of motion within matter and defn
ing the will as the transcendental cause of all motion, reintroduces
the inside/outside strcture that was also adopted at the beginning
of the description of judgent. Just as the vcar begins by sounding
lie an orthodox disciple of Locke when he asserts the truth-value of
sensation, he sounds like an orthodox disciple of Fenelon in asserting
the need to postulate a transcendental prmum mobil. The tran
scendental outside structure becomes productive however (and po
tentially aberrant) only when it is conceived as a polarity rather than
a mere positional relationship, that is, from the moment that a
principle of articulation connects an inside with this outside in a way
that allows for the exchange of properties. In the midst of so many
borrowed arguments and philosophical commonplaces about the
transcendental will , the specifcit of Rousseau's thought manifests
itself in the sudden refusal to grant intelligibility to the principle of
articulation on which the possibility of understanding depends: "It is
no more possible for me to conceive how my will moves my body
than how my sensations afect my soul . I do not even know why one
of those mysteries has seemed more explainable than the other. As
for me, either when I am passive or when I am active, the means of
uniting the two substances seems completely incomprehensible to
me. It is strange that others take that ver incomprehensibility as
their point of departure for combining the two substances, as though
operations s diferent in nature could be better explained in one
suQect than in two" (576) . Wat are here called incomprehensible
are precisely such notions as analog, resemblance, sympathy, or
even proximity which are the ground of understanding and which
the rationalistic and theistic eighteenth centur was tring to re
claim. Far from clariting the obscure link between will and motion
by means of the apparently more verifable (because easier to ob
sere and to quantif) link between matter and mind (or body and
soul) , the vcar lets the darkness of the former encroach upon the
latter. We kow as little about the . (outside) sensation becoming
236
ROUSSEAU
(inside) afect or consciousness as abut the (inside) wll becoming
(outside) motion. As a matter of fact , the indivdual will is unable
ever to get outside itself and to establish a corresponding principle of
exteriorit: "I kow the will only through the feeling of my own will,
and intelligence [entnment] is not better kown to me" (586) .
This does not imply that judgment is enigmatic because, like the will,
it can be metaphorically represented in the guise of a self; it merely
states that judgment is as enigmatic as the will , thoug the mode of
indeterminac or undecidabilit may var depending on whether one
considers the aporia from a voluntaristic or from an epistemological
point of vew. The patterns of error engendered by the will ar not
the same as those engendered by judgment , since both prouce their
own rferential sstems, and are commonly in error only to the
extent that referentiality is itself their error. The referent produced
specifcally by the will is that of se1od, which is open to the same
deconstructive ambivalence as the mor general maning derived
from judgment . The explicit linkge of judgment with eror thus
extends to the idea of selfood, product of the same metaphorical
illusion of proper meaning as the rlational constructs of judgment .
The combination of judgment wth the will engenders i n tun
the mental activity here called intelligence. But the more we advance
in the degee of conceptualization associated with each of these
terms-the text being set up in such a way that , as we move from
judgment to will to intelligence and fnally to freedom, we are
further and further removed from the critical deconstruction that
established the epistemological ambivalence of judgment at the
outset-the more the apora, still quite clearly thematized in the
analysis of judgment , becomes embedded wthin the texture of the
narrative, to the point of making the text into the dramatization of
its own confusions. Examples abound, especially in the closely lined
discussions ofintelligence, as the systematic assertion of the power to
will , and of freedom. It is in these pages (from 577 to 606) that the
strongest statements in support of Rousseau' s theistic orthoox oc
cur: the awe expressed before the deit as a principle of natural
harmony and order, pietistic statements about the innate virtue of a
divine omnipresence, the defnition of moral conscience as innate:
"There i s thus within our souls an innate prnciple of justice and
virtue by which, in spite of our maims, wejudge our acts and those
of others as good or bad, and it is this principle that I call conscience"
( 598) . One should bear in mind that these assertions occur within a
ALLEGORY OF READI NG (PROFESSION DE FOI)
237
context in which the concepts that are being described are consis
tently equated with acts of judgment (as when it is said, in the above
quotation, that "wejuge our acts and those of others . . . ") which
have been show to be constitutively assciated with aberrant totali
zation. And one should also notice that each of these afrmations of
piety and trust is always coupled, sometimes explicitly, sometimes by
implication, with a statement that puts it in question. For instance,
when the vicar wshes to oppose intelligence to sheer random
chance, he chooses a literar text as a model for analog. Such a text
could never be the contingent result ofinfnitely combining the letters
of the alphabet : "If I were told that printer's te had ben thrown at
random and the letters fallen in such a way as to spell out the entire
Ani, I would not deign to tak one step to investigte that false
hood" (579) . 1 6 The statement comes in fact a lot closer to the sen
tence of Mallarme which seems to be saying the reverse ('amais un
coup de des n'abolira Ie hasard") , if one takes into account that , for
the "author" of the Preface to the Nouvll He1oie, the intentionality
("unite d'intention," 580) of a literar text is so undecidable that no
writer can be certain of his own authorship. If Gd is to be present to
his own creation in the same manner that a writer is present to his
text , this leaves ver little authority to the divine intelligence. The
Pesion d foi can therefore posit itself, so to speak, on both sides
of the position taken by Diderot in the Ltre sur ls aveugls wth
regard to the element of chance involved in the creation o the or
ganic world: the natural world is here much more radically contin
gent , disjointed, and inarticulated than the foating organs Diderot
and la Mettrie1 7 are so fond of evokng, but , on the other hand, the
model of a combinator system wth a ver large number of ele
ments seems equally inadequate as an analog for the workngs of
the mind. The abilit of the mind to set up, by means of acts of
judgment , formally coherent structures is never denied, but the on
tological or epistemological authority of the resulting systems, like
that of texts, escaps determination.
More explicit instances of controlled contradiction also appea
in this section of the Poesion d fi. They are dramatically em-
16. Rousseau seems to have a predilect ion for this example which is repeated
in the Let ter to Franquieres (3: 1 139) , the L e to Voltaire (3: 1071) , wher he politely
substitutes the Her for the A, and in a letter to Vernes of Februar 18, 1 758.
17. Diderot , Lt sur l aveu, Ouve complt, 2: 197; l Mettrie, Ss
tm d'Epicure, Ouve philsophiu d Mr. d l Met ( Brlin, 1775) , p. 260.
2
ROUSSEAU
phasized in the contrast between the harmony of a univrse con
ceived as a teleological system without kown endpoint, 1 8 or the
gorifcation of man as "kng of the earth" ( 582) and on the other
hand the utter miser of the human mind incapable of understand
ing the principles of its own workng and torn apart by the con
tradictions of its condition: "Feeling myself swayed and tor by those
two conficting tendencies, I said to myself, 'No, man is not one; I
both exert my will and fail to exert it ; I feel both enslaved and free; I
see what is good, I love it, and I do evi l, etc. ' '' (583) . The contradic
tion comes fully into vew in the way the vicar conceives of G. Since
he rejects any idea of unmediated reelation, the idea of G is
derived, by analogical extension, from the attributes of human
judgent, and rot of nature (for whenever he considers the possibil
it of patterning the relationship beteen Gd and man on that
between man and nature, or subject and object , he rejects the possi
bilit by means of arguments derived from the distinction between
judgment and sensation) . Therefore divine activt is descrbed by
the same terms that were used to defne judgment : "Acting, compar
ing, and choosing are operations of an active and thinkng being;
therefore that being exists . . . " (578) ; ''The being who has both wll
and power, who is active of himself, who moves the universe and
orders all things, that being, whatever he may be, I call Gd. I add to
that name the ideas of intelligence, power, and will . . ." (581 ) . The
mind of Gd and the mind of man resemble each other; man and
Gd are each other's metaphor. 1 9 Hence the fact that the can be
18. In statements such as these: ". . . I never cease to prceive the intimate
corrspondence by means of which the being that compose the universe come to
each other's rescue" (578) ; " . . . I see that each part [of the machine] is made for the
others, I admire its maker in the details of his work, and I am sue that its parts all
move together for some common end . . . " (57) . It is always added tht this
common end is unow, but this do not detract from the fact that the pattem
rmains teleological , "the harmony and the unison of the universe [l'harmonie et
l'accor du tout]" (580) .
1 9. The manuscript of the Peion d fi is more explicit on this pint than
the fnal version: " . . . I call thi bing G. Does this word allow me to kow the
'essence' of the being he represents. No. . . . I wll never kow him in his being. I
can therefore study him only in his attributes . . . I cannot even clearly conceive of
him by his attributes for how could I conceive of them if not b comparison wth
human facul ties" ( Masson, L "prein dfi du vaire Svyar" d]ean-aqu
Rouseau, pp. 145-6) . Se also the note "Deu est intelligent ; mais comment l'est-il?
L' homme" ( ibid. , p. 146) .
ALLEGORY OF READI NG (PROFESSION DE FOI) 239
substituted at will: G can be said, for example, to want man's
good in his stead, and man's freedom to err is accounted for as his
opportunit to prove himself equal to a divine principle that shares
this freedom wi th him (587) . Hence also that the relationship be
tween man and G can be called love: "I worship [ar] this
supreme power and I am touched by his kndness. . . . It is not a
natural consequence of our self-love to honor what protects us and to
love what wants our good? ( 583) . Like the wild metaphor "giant"
which, in the Es y on the Orgn ofLan
g
e, becomes "man, " the
spontaneous metaphor "Gd" can then be institutionalized and
quantifed into a contractual relationship in which God is sid to owe
something to man, and to pay him for the price of labors accom
plished in his behalf. The pro- or regression from love to economic
dependence is a constant characteristic of all moral or social systems
based on the authority of noncontested metaphorical sstems.
On the other hand, however, this Gd who is so much like us
tus out to be as completely alien, unowable, and "outside" as a
pure sensation is before it has been organized by judgment : ". . . as
soon as I tr to contemplate [G] in himself, as soon as I tr to
discover where he is, what he is, what his substance is, he eludes me
and my troubled mind no longer perceives anything" ( 581 ) . Wen
combined with all the prevous passages on the indeterminac of
judgment and on its irresistible tendenc to see seductive similarities
where the do not exist , this passage is not at all similar to the knd of
theological humilit one would fnd, for example, in Malebranche.
The manner in which God is said to be incomprehensible in his
relation to his own being is precisely the same as that in which he is
made all too comprehensible in his relation to man; the myster of
the parousia is not compatible with that of a divne presence show
ing itself provdentially concerned as the voice of an individual moral
conscience. One is reminded of Knt in the Critu ofJugent: "If
one can call a mere representation a cognition, then all our knowl
edge of God is symbolical and whoever schematizes this knowledge
b means of properties such as judgment , will , etc. , which demon
strate only the objective realit of earthly creatures, lapses into an
thropomorhism, just as he lapses into deism if he omits all ele
ments of intuition. "20
The logical pattern of these developments is always the sme
20. Knt , Kritik dr UrtUkraf (Stuttgr, 192) , 59, p. 213.
240
ROUSSEAU
and repeats the aporia of judgment . A concept such as judgment ,
will , or freedom operates deconstructively as a principle uf difer
entiation but then, because of the referentialit inherent in the lin
guistic moel , reintegates by an act of the mind what it had taken
apart on the level of intuition. The correlative of this second opera
tion, regardless of whether it be called the meaning of a judgment ,
the sl that wlls, or the God that freely invents, recovers in its turn
the attributes pf (natural) existence, and can therefore again be
deconstructed by the sme system. The orginal metaphor is shown
to be based on a misleading assumption of identit, but the utter
ance of this negative insight is itself a new metaphor that engenders
its own semantic correlative, its own proper meaning: we move from
sensation to judgment , for example, or from nature to Gd, but what
appears to be a hypostasis is in fact even more vulnerable, logically
speakng, than the entit it claims to supersede. A system ofthis te
is bound to produce bewldering patterns of valorization.
The text of the Profesion d foi becomes indeed increasingly
saturated wth value judgments. I starts out (570 f. ) in a relatively
detached and analyical mode but , as the argument progresses, it
modulates towards highly theatrical oratorical efects . There is a
corresponding shif from the relatively "cold" values of truth and
falsehood (which become values only because of the possibility of
error) to the more turgid values of good and evil. The dynamic
emphasis was present from the start , since judgment is consistently
being described as an at and as a power [puissance] . From the
moment , however, that one reaches such metaphors as the will that
allow for the localization of the active prnciple wthin particular
entities (such as the self, the tension is bound to increase: the in
comprehensibility of the link between will and motion prob
lematizes the relationship between the intent and the direction ofthe
movement , and the solipsistic immanence of the wll ("I know the
will only through the feeling of my ow will . . . ") threatens the ver
possibility of motion with paralysis. The binar inside/outside pat
tern, which could be considered as a merely spatial organization
without implyng valorzation, is now activated by a play of resis
tances and impulses; from being epistemological, the language ofthe
Pofesion d foi becomes ethical . But since the "prime mover" of
judgment is aberrant and represents as a deliberate movement what
is, in fact , a suspended inabilit to know whether or where it should
go (and is thus, in truth, neither "prime" nor "mover") , the system of
ALLEGORY OF READI NG (PROFESSION DE FOI) 21
valorization that lin the modalities of judgent to the values of the
will can never be consistent. Aberrations of moral judgent are a
consequence of epistemological and rhetorical indeterminations.
The Pofesion d foi acts out this inconsistent transformation of
structures into values. The referentiality inherent in judgment be
comes more and more manifest and moves the text closer and closer
to a world of practical reason which will fnally end up in political
realities. The theistic orthodox always associated the structues of
inwardness and exteriority wth the values of good and evil by linkng
inside wth good and outside with evil. The positive valoriztion of
inwardness is part of the historical tradition of pietism out of which
Russeau is witing. One remembers the vicar's exhortation: "Let
us look into ourselves [Rtn e nou-mm] , my young
friend . . . . " Inwardness is the metaphor of virtue, and vice versa.
But the text fails to respect the necessit of this linkge and it crosses
and uncrosses the system established by the inside/outside and the
good/evil polarities at wll. We are told, for instance, that "As I
meditated on the nature of man, I seemed to discover two distinct
principles in him. The frst elevated him to the study of eternal
truths , to love of justice and moral beauty, to those realms of the
intellectual world that the wse delight in contemplating. The second
drew him downward int himel subjected him to the power of his
senses and the passions that are their ministers, and counteracted,
through them, everthing inspired in him by the fst principle" (583,
my italics). More explici tly still , a few pages later, evil is again di
rectly associated wth inwardness and, by implication, the love of
beauty and of virtue directed towards the outside: "If you tae away
from ou hearts that love of what is beautiful , you will tae away all
the charm of life. A man whose vile passions have stifed those
exquisite feelings in his narrow soul , and who, by focussing all his
attention within himel [au dedans de lui] , has come to love no one
but himself, is no longer enraptured by anything, . . . such a
wretched man no longer feels or lives; he is already dead" (596, my
italics). Yet, separated from this passage by only the one paragaph,
the followng statement again reverses the pattern: "The wicked man
fears and avoids himself; he lifs his spirits by going outi himself;
he lok around him with anet, seekng something that wll divert
him . . . derisive laughter is his only pleasure. The virtuous man,
however, fnds only joy, and the source of that joy i s i n himel he is
cheerful alone as in company; he does not draw his contentment
Z2
ROUSSEAU
from those around him: he communicates it t o them" (597, my
italics) . One can argue, of course, that , psychologically speaking,
there canjust as well be a "bad" as a "good" inwardness. But this has
little bearing upon a text that is not set up along psychological lines
but structured by means of a diferential inside/outside axis; in such
a text , the values associated wth these two dimensions will necessar
ily carr a decisive exegetic weight.
The occurrence of such systematic value-displacements thus ac
quires central importance in the reading of the Pofsion dfoi. They
are not simply chiasmic reversals allowing the (inside) good to be
called "evil" and the (outside) evil to b called "good. " The system is
not smetrical , since it postulated, from the start , the nonidentity
of inside and outside, the "supplementarity" of the one with regard
to the other. On the level of judgment , the asymetr leads to the
play of contradictions and paradoxes, the logical tensions that have
eared Rousseau the frequent accusation of sophistr. Deconstructive
readings can point out the unwarranted identifcations achieved by
substitution, but they are powerless to prevent their recurrence even
in their own discourse, and to uncross, so to speak, the aberrant
exchanges that have taken place. Their gesture merely reiterates the
rhetorical defguration that caused the error in the frst place. They
leave a margin of error, a residue of logical tension that prevents the
closure of the deconstructive discourse and accounts for its narrative
and allegorcal mode. When this process is described in terms of will
or freedom and thus transferred to the level of reference, the difer
ential residue is bound to become manifest as an empirical aware
ness that afct and indeed constitutes a world in which it now
appears to be "takng place"; a mind, a consciousness, a self. The
abstract attributes of truth and falsehood grow more and more
concrete and fnd themselves a place, a stage on which to act out the
temporal sequence of their occurrence. Judgment becomes a "spec
tacle" ( 596) , a pathetic action that afects us like a theatrical repre
sentation. Trng to persuade his interlocutor of the qualit of vrtue,
the vcar resorts as by instinct to analogies from the theatre: "When
you see a play, which characters win your sypathy?" (596) , he asks.
As the confusion between structure and value increases, the tone and
the terminolog of the text glide almost imperceptibly from the lan
guage of judgment to the language of the afections, and judgment
fnally openly declares itselfto be another name for "sentiment" (still
distinct , at this point , from "sensation," againt which, it wll be
ALLEGORY OF READI NG (PROFESSION DE FOI) 243
remembered, judgment was originally defned) . The ambiguously
valorized "inner" world of consciousness, of which it can no longer
be said whether it is the seat of good or evi l , bcomes the afective
space engendered by this ethical indecisiveness: "The acts of our
conscience are not judgments, but feelings. Athough all our ideas
come from outside us, the feelings that evaluate them are inside
ourselves , and it is solely by those feelings that we know the ftness or
unftness that exists between us and those things which we must
respct or shun" (599) .
The principle of valorization according to which this evaluation
can be carried out has itself to become increasingy literal . The
eudaemonic polarity of pleasure and pain replaces the moral polarity
of good and evl . The text behaves as if, at this point (599) , the
question that had opened up its entire inquir had been decisively
answered. We are now supposed to kow the answer to the critical
"But who am I?" (570) that was to sanction the recourse to imma
nent evdence in the understanding of our being-in-the-world. What
has in fact been established is the gradual loss of authority of any
immanent judgment or any immanent value whatsoever. At the
same t ime, and by means of the same argument , the alternative
recourse to a transcendental source of authority, such as nature, or
God, has also been defnitively foreclosed. The aporia of truth and
falsehood has turned into the confusion of good and evl and ended
up in an entirely arbitrar valorization in terms of pain and pleasure.
Virtue becomes fnally justifed in terms of an erotic pleasure princi
ple, a moral libido that seems not easily compatible with the piety of
the inner voice of conscience but that consistently acts out the rhetor
ical system of the text .
The turn towards eudaemonic valorization is more apparent ,
for obvous reasons, i n the so called Ltre morls addressed by
Rousseau to Sophie d'Houdetot , a text that is ancillar to the Pf
sin dfoi and that dates from 1 757. In this text , vrtue is spoken ofin
terms of a narcissistic economy of personal well-being, accessible
only to those who can aford a great deal of leisure: "Ever month,
take a span of two or three days away from your pleasures and you
business, and devote it to the most important task of all . . . . I don't
expect you to concentrate from the start on profound meditations, I
only insist that you keep your soul in a state of quiet languor that will
allow it to tum inward upon itselfand to exclude whatever is alien to
its own being. In that state, you may ask, what wll I do? Nothing at
ROUSSEAU
all . Just let natural unrest take over; i n solitude, it wl l not take long
to manifest itself, whether we want it or not . . . . Just as one re
warms a numbed part of the body by rubbing it delicately, so the
soul , numbed by a long state of inaction, revives in the gentle
warmth ofa well-tempered motion . . . . One must remind it of the
afections that have caused it pleasure, but not by ways of the senses;
it must happen by the proper feelings and by intellectual pleas
ures . . . . Watever state the soul may be in, a feeling of pleasure at
doing the right thing remains and seres as a foothold for all other
virtues. It is by cultivating this feeling that one comes to love and to
fnd pleasure in one's own company ( 1 1 1 5-16) . The language of
self-seduction is less obvious in the mouth of the vicar, but he is
saying nothing else when virtue is called "the charm of life" (596) or
"the pure voluptuousness born from self-satisfaction" (591) .
The point is not that the foundation of moral judgment in the
pleasure principle is in any way ethically or psychologically wrong.
Such returns to the physiological foundations of the notions of right
and wong are not at al l surprising in the centur of morl senitive,
nor for that matter in any other centu; without even having re
course to Freud, we can refer, for instance, to the importance given
by Nietzsche to the pleasure/pain polarity in his critique of
metaphysical concepts. 21 Wat matters here is that the reintroduc
tion of these notions, at this point in the Prfsin d joi, illustrates
the viciously circular (in the Nietzschean sense) structure of Rous
seau's theological discourse. For he has no illusions about the consis
tenc of eudaemonic systems of valorization. The association of vir
tue wth pleasure (as in the text from the Ltre mr1 that has just
ben quoted) can at once be reversed and the self-love turn into the
amour propre which is at the base of bad faith and of bad judgment:
"The practice of virtue naturally fatters our vanit [amour prpre] by
an idea of superiorit. We remember goo deeds as proof that we
have the strength to satisf even the needs of others afer our own
needs have been fulflled. This feeling of power [puissance] makes life
more pleasant and maes us cohabitate more easily with ourselves"
( 1 1 16) . But the ambivalence of this will-to-power morality is not the
main complication of our text . The sliding pro- or regession from
judgment to feeling, from epistemolog to eudaemony, a motion
which takes place entirely wthin the conceptual system that consti-
21 . For exmple in Fredrch Nietzsche, Th Will to Power, translated by Wal
ter Kufmann and R. J. Hollingdale ( New York, 196) , pp. 371-74, sctions 699-703.
ALLEGORY OF READI NG (PROFESSION DE FOI) 25
tutes the tet ofthe Pfesin d foi and is therefore accounted for, at
all times, by its logic, requires, as a thematic assertion, the necessa
reintroduction of the rhetorcal strctue that was explicitly banned
at the stat . The serpent bites its own tail . The vicar has to reafm,
at the end of his argument , the prority of the categor against which
his argument has been consistently directed. He has to restate his
belief in the metaphorcal analog between mind and nature: ". . .
we feel before we kow, and just as we do not learn to will ou own
goo and avoid what is harmful to us, but receive that will from
nature, love of good and hatred of evil are as natural to us as self
love" ( 599) . The equation of will with nature is precisely what the
vicar's judgment persistently puts in question. Within the context of
this deconstruction, the fnal part of this quotation can be read
ironically, since nothing could b more problematic than the nat
ralness of self-love, a passion which, like all passions, merely repeats
the aberrations of fgural language. Nothing therefore prevents the
deconstructive labor that has brought us to this point from starting
all over again.
The naive historical question from which we started out
should the Pfsin d fi be called a theistic text?-must remain
unanswerable. The text both is and is not the theistic document it is
assumed to b. It is not the simple negation of the faith it seems to
proclaim, since it ends up by accounting in a manner that cannot be
refuted for the necessr occurrence of this faith. But it also de
nounces it as aberrant . A text such as the Pfsin d foi can
literally be called "unreadable" in that it leads to a set of assertions
that radically exclude each other. Nor are these assertions mere neu
tral constations; the are exhortative perormatives that require
the passage from sheer enunciation to action. They compel us to
choose while destroyi ng the foundations of any choice. The tell the
allegor of a judicial decision that can be neither judicious nor just .
As in the plays of Keist , the verdict repeats the crme it condemns. If,
afer reading the Pfsin d foi, we are tempted to convert our
selves to "theism," we stand convicted of foolishness in the court of
the intellect. But if we decide that belief, in the most extensive use of
the term (which must include all possible forms of idolatr and
ideolog) can once and forever b overcome by the enlightened
mind, then this twilight of the idols will b all the more foolish in not
recognizing itself as the frst vctim of its occurence. One sees from
this that the impossibilit of reading should not be taen too lightly.
11 Pomises
(Sl Contrat)
THE CONNECTION BETWEEN THE POLITICAL AND THE
religous writg of Russeau i enigatc ad, at fst sigt , ently
contradictor. Russeau's theolog and his political theor seem to
lead in opposite directions. There have been excellent book witten
abut Russeau's political theor that don't even mention his reli
gous concerns, and vice versa. I Yet the second par of the Nouvll
Helie combines the discussion of political institutions with theolog
ical considerations and at least sugests a close, albeit unformulated
interrelationship between both. Ad the &l Conta, which obvi
ously proposes a model for political institutions and refects on the
authorit of legal langage, has to reintroduce religious themes at
at least one crucial point. 2 The difcult may well stem from the use
of such thematic terms as "political" or "religous," as if their refer
ential status were clearly established and could b understoo with
out regard for the rhetorical mode of their utilization.
1 . E.g. , Mario Einaudi, Th Early Rusu (Ithaca, N.Y. , 1967) or, in the other
dirtion, Luis Athussr, "Su Ie Cnt S (Ls Dcalages)" in L'imp dJ.
J. Rus, Cairs pour l'anlyse, vol. 8 (Paris, 1970) . pp. 5-. The problem ofthe
rlationship btween Russau's theolog and his political theores is, of coure, o
central imprtance in all historcal and themtic studies of Russau's work a a
whole, for example, in the book and aticles of Rbr Drathe, Pere Bugelin,
Henr Guhier, )ean Starobinski , Rlnd Grmsley and others. The notes to the thid
volume of the Pleiade edition (Frt plituJ refect the preent state of the
question in academic Russau interretation. The polemical implications of the
reading that is here bing sugested have not ben develop, since the emphasis
falls on the theoretical stat of misreading in general rather than on actual (mis)
reading of their histor.
2. Particularly in the section on the lawgver (Sl Cnta, bok 2, chapter
7, pp. 31 f. ) and in the pnultimte section of the text (Sl Cont, bok 4,
chpter 8, pp. 46-69) . Al quotations a from). ). Rusu, Ouv cmpl, e.
Brar Ggnebin and Marcel Ryond (Pa: Gallmard [Bibliotheque de l
Pleiade] , 19), vol. 3. Transltions are my ow or based on the version establishe
by Lowell Bair i Th Etal Rusu (Ne York: Rndom Hous, 1974) with
ocasional moifcations.
PROMI SES (SOCIAL CONTRCT) 247
It may b just as difcult to decide upon the rhetorical status of
theoret ical texts such as the Prfsion d fi and the Scial Contat
as on a fction such as the Nouvell Heli. The diference beteen a
fctional and a theoretical text carries ver little weight in the case of
Rousseau. By reading the unreadabilit of the Prfsin d foi, we
found it to be structured exactly like the Nouvell Heloie: the decon
struction of a metaphorical moel (called "love" in the Nouvell
Heloie, 'judgment" in the Prfesin ) leads to its replacement by
homological text systems whose referential authorit is both asserted
and undermined by their fgral logic. The resulting "meanings" can
be said to be ethical , religious, or eudaemonic, but each of these
thematic categories is torn apart by the aporia that constitutes i t ,
thus makng the categories efective to the precise extent that they
eliminate the value system in which their classifcation is gounded.
I fwe choose to call this pattern an allegor of unread abilit or simply
an allegor, then it should b clea that the Pfsion d foi, like Juli,
is an allegor and that no distinction can be made beteen both texts
from the point of view of a genre theor based on rhetorical models.
The fact that one narrates concepts whereas the other narrates some
thing called characters is irrelevant from a rhetorical perspective.
But if the Profesion d foi is an allegor of (non) signifcation,
can the same still be said of the Sil Contrat? Again, no reliable
answer can be given by merely quoting or paraphrasing the text
wi thout reading i t . And to read the Scial Contrct is, for instance
and among other things, to determine the relationship between gen
eral will and particular will , two notions that obvously play a pre
dominant part in the organization of the text .
A frst difcult in the use of the polarit between the general
and the particular will is lexicological and stems from the apparently
interchangeable use of the terms "natural" (as in relign nturell,
drit naturel, etc. ) and "particular" or "indivdual ," both used in
opposition to "civil" or "collective. " Rousseau follows common usage
in speakng of natural law, natural religion, or natural freedom ( p.
293) ; he does not use "volonte naturelle" however, but would rather
have chosen, in opposition to "volonte generale," the term "volonte
particuliere. "3 Yet , taken literally, "particular" is clearly not the same
3. The specifc expression "volonte particuliere" occus rarly or not at all in
this for, but Russeau spa frequently of"fait ," "doit" Cpp. 306, 307) , "objt" C p.
378) or "acte" C p. 287) "particulier" in a way that leaves no doubt that what is in
conflict whn "man" is opposed to "citien" are the categores of particularty and
24
ROUSSEAU
as "natural" ; i f we say, for example, that the frst part of the Nouvell
He101 deals with the particular, or individual, relationships be
tween julie and Saint-Peux in contrast to the second part which, at
least at times, deals with public, collect ive relationships between the
inhabitants of Clarens, i t does not follow that , in the frst three
book, julie and Saint-Peux are in the state of nature as the term is
used in the Son Diourse. A certain amount of confusion results
from Russeau's interchangeable use of "natural" and "particular,"
especially since his sense of the complexities of selfood puts the
indivduals he portrays far beond the simplicit of the state of na
ture. This is true of fctional entities such as the "characters" of the
Nouvell Heloie (if one wshes to consider them as such) as well as of
actual human beings, including Rousseau himself, in the autobio
graphical wtings. It would b absurd, for instance, to consider the
Confesin as more "natural" than the Scil Contrat because it
deals with indivdual experiences rather than wth societies. The case
of Emil is somewhat diferent, since the diegetic narrative is sup
posed to follow the histor of an empirical human being from the
start and along chronological lines. 4 This forces upon us the contrast
between a "natural" child and a corupted citizen, an anti thetical
pattern of innocence and exprience. The rhetorical mode of Emile
produces the opposition between nature and society as a textual
necessit. No such polarity functions in the conssion, since Rous
seau never claims to narrate anything about the child jean:acques
that is not directly remembered by him. He is thus at least twce
removed from the preconscious condition of nature: the experiences
of a highly self-conscious and "denatur" child are told by the dis
fgred fgre of a highly self-conscious narrator.
The lexicological confusion between "natural" and "particular"
thus has only limited theoretical interest , although it certainl has
been responsible for many aberrations in the interpretation of Rous
seau. It nevertheless provides a point of entrance into the remarkably
smooth and homogeneous textual surface of the Scil Contrat. For
it agin attracts attention to the danger of hyostatizing such con
cepts as "nature," "individual" or "societ as if they were the desig-
gneralit. The same polarity opposes private to public in such expressions as "per
sonne publique" and "personne privee. "
4. A problem that does not aise in the Son Dicure where the natural
origin of mankind is a fction and the diachrony of the narative only exsts on the
level of the signifer.
PROMI SES ( SOCIAL CONTRCT) 249
nation of substantial entities. Russeau can legitimately shif these
terms around and confuse the names of two such divergent semantic
felds as those covered by "nature" and "particularity" because the
designate relational properties, patterns of relational integation or
disintegation, and not units or modes of being. This may be easier
to admit in the case of superstuctures such as civl societ, or the arts,
or technolog, but it pertains to the term natre as well. Russeau
calls natal any stage of relational integration that precedes in de
gree the stage presently under examination. In the analysis of con
ceptualization, the "natural" stage that precedes the concept is de
nomination; in the analysis of metaphor, the natual fgure would be
metonymy; in the critique of judgment , it is sensation or perception;
in the case of generality, any prevous mode of particularit, etc. The
deconstruction of a system of relationships always reveals a more
fragmented stage that can be called natual wth regard to the sys
tem that is being undone. Because it also functions as the negtive
truth of the deconstructive process, the "natural" pattern authorita
tively substitutes its relational system for the one it helped to dissolve.
In so doing, it conceals the fact that it is itself one system of relations
among others, and it presents itself as the sole and tre order of
things, as nature and not as structure. But since a deconstrction
always has for its target to reveal the exstence of hidden articula
tions and fragmentations wthin assumedly monadic totalities, na
ture turns out to b a self-deconstructive term. It engenders endless
other "natures" in an eternally repeated pattern of regession. Nature
deconstructs nature, hence the ambiguous valorization of the term
throughout Rousseau's work. Far from denoting a homogeneous
mode of being, "nature" connotes a process of deconstruction redou
bled by its ow fallacious retotalization. In the oppsition bteen
private and public, or particular and general, the frst term is the
"natural" counterpart of the second, provided one reads "natual" as
has just been sugested. We conclude that there is no structural
diference between the couple linkng "volonte particuliere" to ''vo
lonte generale" and, on the other hand, such pairng as "doit" or
"religion nature lie" "de l'homme" with "droit" or "religion civle"
"du citoyen. "
Any Rousseau text that puts such pol.arities into play wll there
fore have to set up the fction of a natural process that functions both
as a deconstructive instrument and as the outcome of the decon
struct ion. Frequently enough, the fction is provded by a cont empo-
250
ROUSSEAU
rar or traditional text written by someone else: i n the Pofsin d
foi, it was primarily Helvetius and Quesnay that were thus being
used; in the frst version of the Scil Contat, Diderot's Enclpe
entr on "Doit naturel" funishes the appropriate target. Lacking a
suitable formulation, Rousseau sets one up himself, thus conveying
an impression of self-contradiction that has considerably enriched, if
not clarifed, the histor of his interpretation. This is in part what
happens in the Son Dioure, where neither Hobbes, nor Condillac,
nor any of the other polemical opponents, provides an adequate
natural model and where Russeau therefore has to i nvent one him
self. A similar, somewhat more complex instance of the same strateg
occus in a tet that has close afnities with some aspects of the Scial
Contat, the fragment that the editors of the Pleiade edition have
included under the title "D bonheu public. "
Althoug the fragmentar state of "D bonheur public" (as wel
as the fact that it originated as an occasional improvisation in reply
to a questionnaie sent out by the Societe economique de Berne)
makes a sustained reading difcult , the notes nevertheless illustrate
the odd logical shape of Rousseau's political discourse. And because
it deals with the opposition between private and public values as they
relate to the political constitution of the State, the brief text is like a
blueprint for the more elaborate structue that supports the Scil
Contat. In trth, "Du bonheur public" is not based on a dialectic of
private as opposed to public or social identity; it considers the possi
bilit of a readable semiolog of private happiness that would be
based on analogies between inside feeling and their outside man
ifestations only in orer to reject it out of hand: "Happiness is not
pleasure; it is not a feeting stiring of the soul, but a permanent and
entirely inward feeling, that can only be evaluate by the person who
experiences it. No one can therefore decide with certaint if someone
else is happy, nor can he, as a result of this, come to know with
certaint the sigs that bear witness to the happiness of individuals"
(p. 510, my italics) . 5 Consequently, there can be no easy metaphori
cal totalization fom personal to social well-being, based on an
analogical resemblance between both: "It is . . . not by the feeling
that the citizens have of their happiness, not consequently by their
happiness itself that one can judge the prosperit of the State" (p.
5. A little futher in the text , Russeau also spea of " . . . the tre sig that
might characterize the well-bing of a people" (p. 512, my italics) .
PROMI SES (SOCIAL CONTRACT) 251
513) , or: "IfI had deduced the idea of happiness collectively from the
particular state of happiness of ever citizen that makes up the State,
I could have said something that would have been easier to gasp for
many readers, but aside from the fact that no conclusions could ever
have been drawn from these metaphysical notions that depend on
the way of thinng, the mood and the disposition of each individual ,
I would have gven a ver inaccurate defnition" (p. 512) . The text
takes the deconstruction of a private inwardness that could be
equated with a natural inwardness for granted; the spontaneous
manifestations of the happy consciousness are so unnatural as to be
entirely unfathomable and beyond obseration. The actual evocation
of private happiness, s freely developed in many other Russeau
texts such as the Conesin or the Re, does not even require
formulation here. Yet it operates as a totalizing power based on
natural properties, and it at once replaces the dismissed, natural
afectivt of the individual subject b a natural afectivit of the
goup that can be interpreted precisely as the self has just been
show nt to be. Natual societies in which "men [will be] social
[civil] by nature . . . " (p. 510) are assumed to exist or (what
amounts to the same thing) to have existed or to be conceivable in
the future. In such societies, the semantics of afectivity, opaque in
the case of individuals, are transparent and reliable: "[The] virtues
[and the] vices [of political societiesJ are all apparent and visible,
their inner feeling is a public one. . . . For any ee that can see, they
are what the seem to be, and one can safely evaluate Uuger] their
moral being" (p. 510) . This being granted, the totalization is bound
to ensue without further delay: "[men] will be united, they will be
virtuous, they wll be happy, and their felicit will be the well-being
of the Republic; for since they receive all their being from the State,
they owe everhing to the State. The Rpublic will ow all they own
and will be all they are" ("car n'etant rien que par la republique ils
ne seront rien que pour elle, elle aura tout ce qu'ils ont et sera tout ce
qu' ils sont") . Such sentences automatically fall back into the familiar
diction of "all" and "nothing," or "all" and "one," in the reconcilia
tion "schon langst Eines und Ales genannt" (Holderlin) of the
pantheistic h ki pan. The model for this utopia is the reconcilia
tion of the most natural of goups, the family, with the State: "The
family, pointing to its children, will say: it is in them that I fourish"
(p. 51 1 ) . It also reconciles moral virtue with economic wealth and
makes property innocent by makng it collective; the word "bien" can
252
ROUSSEAU
be used in its ethical register (as the opposite of "mal") as well as in
the economic sense of "real" estate. That , in the sentence which
speaks of "les biens, les maux . . . " of political societies (p. 510) ,
"biens" also means wealth is clear from its use a few lines further on:
"au tnsor public, vous aurezjoint les bin des particuliers" (p. 511,
my i talics) .
As it denies the validit ofthe metaphor that unites the self wth
societ, the text, by the same token, elaborates a new metaphor, the
"natural" plitical societ or family which, in its turn, fulflls the
totalization that was denied to the frst binar pair. But the logic of
totalization works both ways, towar the on as well as toward the
all , and the welfare of the natural society is bound to restore the
well-being of the individual who relates to it as part relates to whole
in an organic syecdoche, as the "member" relates to the ''body'' in
the political unit. Therefore Rousseau has to state the ver opposite of
his initial assertion: ". . . do not imagine that the State could be
happy when all its members are in distress. This ethical fction that
you call public happiness is, by itself, a chimera: if the feeling of
well-being is not felt by anyone, it has no existence and no family can
fourish if its children do not prosper" (p. 510) . This is stated in the
paragraph immediately followng the one which denied the sig
nifcance of any indivdual well-being for the societ as a whole, and
it wll be followed, not much further, by the equally categorical
assertion: " . . . [that it is] therefore not by [ the] happiness [of the
citizens] that one can measue the prosprit of the State. "
The occurrence of such contradictions wthin the confnes of a
few lines obviously does not have the same efect in a tentative,
unfnished and disjointed text as in a more continuous argument. On
the other hand, one may well wonder, with equally good reason,
whether the pattern of contradiction in this fragmented composition
does not represent a more faithful outline of Rousseau' s thought
patterns, simply because the narrative developments and transitions
that conceal incompatible afrmations merely by putting some
space between them are lackng in this case. A text such as this one
bears a close resemblance to what is generally referred to, rather
inaccurately, as Nietzsche's aphoristic manner, as we know it from
Human all to humn on. This discontinuous format goes back, in
Rousseau, at least as far as the notes to the Scon Dicourse and
represents probably the most characteristic dimension of his stle.
The reading of"Du boneur public" is not completed when the
PROMI SES (SOCIAL CONTRCT) 25
fgural pattern of contradiction has been pointed out. Short as it is,
the text contains the elements necessr to the second deconstrction
that can be gafted onto the frst undoing ofthe "natural" metaphor,
and that raises its fgural status to the second power, mang it into
the fgral deconstrction of the prior deconstruction of a fge. For
the text also states, this time truly in the for of an aphorism,
without futher context : "The moral condition of a people is less the
result of the absolute condition of its members than of the relation
ships among them" (p. 511) . One is reminded of the strcture of
judgment as a posited relationship (in the Pofsion d joi ) and also
of the statement about love in the Nouvll Helf: " . . . the source of
happiness does not reside wholly i the desied objct nor i the heat
that possesses i t , but in the relationship of the one to the other"
(Oeuve complet, 2:225) . An entirely diferent principle of organi
zation is introduced by this description. If the principle of collectivi
zation or generalization that constitutes what is here called a
"people" does not operate between part and whole but is determined
by the rlationship that the diferent parts, as parts, establish be
tween each other, then the rhetorical structure is no longer the same
as in binar structures. The principle of diferentiation no longer
oprates between to entities whose diference is both redundant
(since it is posited from the start) and transcended (since it is sus
pended at the end) ; it now operates to reveal diferences where a
metaphorical totalization had created the illusion of an identity, a
delusive generalit in such wors as "man," "self," "people," or
"State," all of which sugest that, to the extent that the are men, or
people, or States, all men, people, and States are essentially the same.
Groups constituted on the basis of relationships which no longer
claim to be natural engender diferent systems of interaction, in
relation to themselves as well as to other groups or entities. Since the
principle that establishes their genrl character as goup is no
longer a principle of necessit, but the result of an uncertain act of
judgment ratifed by convention, it follows that the principle of
generalization that constituted it is by no means unique. The sme
entit can thus be inscribed wthin diverse sytems that are not
necessarily compatible. The can be considered from diferent points
of view without necessarly allowing for a coordination ofthese var
ous perspectives . Neither does their interference wth other systems
necessarly allow for specular exchanges or integrations.
The shift from a (deconstructed) binar moel to this still un-
25
ROUSSEAU
identifed "other" model occurs i n "D bonheur public" when Rous
seau abruptly changes goud, abandons the bina model that seek
to derive political well-being from private happiness altogether, and
afrms, as a "ver general idea . . . to which no reasonable ma
could, I believe , refuse his assent" (p. 512) that " . . . the happiest
nation is the one that can most easily dispense with all others and the
most flourishing the one the others can least dispense with" ("la
nation la plus heureuse est celIe qui peut Ie plus aisement se passer
de toutes les autres, et la plus florissante est celIe dont les autres
peuvent Ie moins se passer" [po 512]) . The langage shifs from the
qualitative and unfathomable "will to happiness" to an outspoken
will to power quantifed in terms of economic and milita interests.
The decisive relationship is no longer between constituting and con
stituted elements. Within a political entity, no necessar lin con
nects individual to collective well-being; to pursue the problem of
this relationship (as "D bonheur public" set out by doing) is to
pursue a false problem. It is not irrelevant to raise the question ofthe
"happiness" of a political entity such as a State, but it can only be
considered in terms of the relationship of one State to another. The
ver concept of a political entity, be it a State, a class or a person, also
changes: an entit can be called political , not because it is collective
(constituted by a plurality of similar units) , but precisely because it is
not , because it sets up relationships with other entities on a non
constitutive basis. The encounter between one political unit and
another is not a generalization in which a structue is extended on
the basis of a principle of similarity (or of a proximity considered as
similarit) to include both under its common aegis. Just as the unit
itselfis not the outcome of such a generaliztion, the relationships of
the units among each other are not stated in terms of afnities,
analogies, common properties or any other principle of metaphorical
exchange. They depend instead on the ability of one entit, regadess
of similarities, to keep the relationship to another contingent, "to b
able to dispense with all other [nations] . " Ifthis degee of autachy is
achieved, relationships with other States are still possible, and
perhaps desirable, but the are no longer compulsor. To the extent
that a less fortunate State is unable to achieve this and remains
dependent for its exstence on necessar link, it is not a truly politi
cal entit, not really a State at all. In other words, the strcture
postulates the necessar exstence of radical estrangement between
political enti ties. Autarchy, as it is here conceived, is not a principle
of autonomy, still less of totalization. The accent does not fall on
PROMI SES (SOCIAL CONTRACT) 255
freedom as a positive force but on the ability to dispnse with others;
the worst that can happen politically is having to have recourse to
"strangers. " Such patterns of estrangement are an inevitable aspect
of political structures; it is well known, for instance, that Rousseau
refused to believe in perpetual peace and that, by consistently ar
g

ing against Hobbes that the state of war is not a natural state, he had
to see war as a necessar moment in the political process: " . . . far
from being natural to man, the state of war is born from peace, or at
the least from the precautions men had to take in order to insure
themselves of a durable peace. "6
In its stress on separation and solitude, on the fragmented dif
ferentiation of entities rather than on their unity, the condition of
political estrangement is reminiscent of the state of nature. Tis is
not surprising, since the fction of a natural "state" results from the
deconstruction of metaphorical patterns based on binar models-
which is exactly how Rousseau's defnition of political happiness here
comes into being, afer the antithetical system that deduces public
from private happiness has been allowed to destroy itself b running
its course. For the same reason, the rediscover of diferential mo
ments, such as those sugested by the term estrangement, also sig
nals the inevitable relapse into patterns of totaliztion. The sentence
that asserts the diferentiation ("the happiest nation is the one that
can most easily dispense with all others") also asserts the simultane
ous reconstruction of an aberrant totality: "the most fourishing is
the one the others can least dispense with. " The shif from qualita
tive happiness [nation hureue] to quantitative prosperity [naton
forsant] is revealing. The synonymy of both terms is asserted as
part ofthe sstem: ". . . if money makes the rich happy, it is less by
its immediate possession than because it enables them, frt to satisf
their needs and to carr out their will in all things, without havng to
depend on anyone, and second to exercise command over others and
keep them in their dependence" (pp. 513-14) . 1 Yet a shif is implied
by the quantifcation, which is also a sureptitious reintrouction of
6. Ecr sur ['Ab d Sint-P e, 3: 61 1 . Sealso: "Hobbs' eror is not to have
established the presence of a state of war among independent men in a societ, but
to have assumed that this is the natural condition of the species, and to have made it
the cause of the vices of which it i the efect" (Sil Cntrat, frst version, p. 28) .
Kant's famous text On Perpetul Peae ( 1795) i s of course highly relevant here.
7. ThE pint is further developd, in terms of a semiolog of money, in the
fragment "L luxe, Ie commerce et les arts" (pp. 520 f. ) . Se also Dr sur l
rhs, not included in the Pleiade edition, published b F. Bovet in 1853.
256
ROUSSEAU
conceptual number metaphors into the deconstructed system. The
"happ State is entirely self-sufcient and does not need to establish
relationships with any other nation, but the ver strength resulting
from this independence allows it, almost in spite of itself, to assert its
power over others and make them in turn aware of their depen
dency: "The securit, the conseration [of the State] require that it
become more powerful than all its neighbors. It can only augment,
nourish and exercise its power at their expense, and althoug it does
not have to look for subsistence byond its boundaries, it nevertheless
always seek for new members that reinforce its own strength [qui
lui donnent une conitance plu inebrnlabl] " (Ecrit sur l 'Abbe d
Sint-Pre, 3:605) . The political power does not remain in its condi
tion of fragmented isolation in which it is satisfed to consider the
other State as a pure stranger. Carried by the metaphorical structure
ofthe number system, it enters into relationships of comparison that
necessarily lead back to totalizations from part to whole: "Since the
size of the political body is entirely relative, it is forced to enter
steadily into comparison in order to know itself; it depends on ever
thing that surrounds it, and must take an interest in everhing that
happens outside, although it would want to remain self-confned,
without gains or losses . . . . Its ver strength, makng its relation
ships to others more constant , makes all its actions further-reaching
and all its quarrels more dangerous" (Ecrit sur l 'Abbe d Saint
Piere, 3:605) . Consciousness of selfood [s connaite] , whether
individual or political , is i tself dependent on a relationship of power
and originates with this relationship. The danger of the situation is
not only the actual damage done to the others, but the reintroduction
of a master/slave relationship of mutual dependenc wthin a system
that had come into being by overcoming the fallacies of this model . s
The fact that the strong and "happy" State comes t o depend on the
dependenc of the other as if it were a necessit dissolves the struc
ture to which it owed its existence. For the master/slave relationship
is not the (non) relationship of pure estrangement that was posited as
the necessar condition for a political entit to come into being. It
clearly is a polarit susceptible of dialectical exchange: master and
slave are no strangers to each other, as little as the conquered State
8. " . . . the war (btween two nations) can only cease when both of them
freev proclaim their renunciation of war. It follows that, as a consequence of the
relationship between master and slave, they continue even in spite of themselves, of
bing in a state of war" (Ert sur l'Abb d Sint-Pre, 3: 615) .
PROMI SES (SOCIAL CONTRACT)
257
remains a stranger to its conqueror. "Du bonheur public," like all the
other allegorical Russeau texts, reintroduces the metaphorical
model whose deconstruction had been the reason for its own elab
oration. It is therefore just as "unreadable" as the other allegories we
have considered.
Does this mean that it can, wthout futher qualifcation, be
equated with such previously considered texts asJuli or the Pfe
sion dfoi? It would then difer from them only thematically, by its
political "content ," but not rhetorically. Ifwe are correct in assuming
that the logical structure of"D bonheur public" also operates as the
organizing principle of the Scil Contat-a point that remains to
be shown-it would follow that the Sl Contat could be called
an allegor of the sme te and for the same reason as the Pfs
sin d foi. This conclusion would, in a sense, make the consideration
of the Scil Contat and of Rousseau's political texts in general
redundant for the rhetorical analysis of his work as a whole. It would
merely confrm what we already know, that Rousseau's fctional as
well as his discursive witings are allegories of (non)sigifcation or
of unreadabilit, allegories of the decontruction and the reitroduc
tion of metaphorical models. We would merely have gained yet
another version of this same insight . This woud not lead to a
refnement of the question that remained precariously suspended at
the end of the reading of the Pofsion d foi, when the referential
power of the allegorical narrative seemed to be at its most efective
when its epistemological authority was most throughly discredited.
The political question would then only have a digressive importance
in the reading of Rousseau; like the metaphor of selfhood, it would
have no privileged interpretative function. Such a conclusion is not
wthout consequence: for example, the naive [impene] distinction
between a "literar and a "political" Rousseau from which we started
out as from the empirical donne provided by the present state of
Russeau studies has, to a considerable extent , been overcome. I t no
longer makes sense to consider Juli as more or less "literar than
the Scil Contat; neither can the assumed inconsistencies and
contradictions of the political theorist b explained away b calling
them "literar" and attributing them to the discrepancies between
"l' analyse intellectuelle" and "l'elan du coeur. "9 One can no longer
9. The exression is b Pere Burgelin in his introduction to the edition of the
Pfsin d foi du vcair savoyar (Ouv com
p
lte, 4:cxiii) .
25
ROUSSEAU
call Rousseau an "admirable" writer and dismiss the political
theorist or, conversely, praise the rational rigor of the political
analyses while writing of the more disquieting aspcts of the imagi
native and autobiographical writings as more or less accidental
patholog. The rhetorical reading leaves these fallacies behind by
accounting, at least to some degee, for their predictable occurrence.
With the assistance of the plitical writings, it may however be
possible to take one futher step.
"D bonheur public" appears to be "privileged" i n at least one
respect : it isolates a model for the elaboration and the comportment
of plitical entities that is more rigorous and systematic than the
models we have encountered up till now. It is true that this model
never asserts itself as an actual state of being (in all the meanings of
the term "state," be it as state of nature, state of war, the State, etc. ) .
It is at once overtaken by other rhetorical patterns, similar in struc
ture to conceptual metaphors. But the question remains what is
being overtaken and how this "relapse" is to be understood. The
Sil Contat, the most complete section of Rousseau's planned
treatise on political institutions, provides information on this point.
We are not here concerned wth the technically plitical signifcance
of this text , still less wth an evaluation of the political and ethical
praxis that can be derived from it. Our reading merely tries to defne
the rhetorical patterns that orgnize the distribution and the move
ment of the ke ten- whie contending that questions of valoriza
tion can be relevantly considered only afer the rhetorical status of
the text has been clarifed.
That the Scil Contrat implies a deconstruction of a binar
metaphorical system similar to that operating between personal and
public happiness in "Du bonheur public" is hardly apparent in the
fnal text but becomes much more noticeable if one takes into ac
count the earlier version kown as the manurt d Gen. lO This
version begns wth a genetic section that sets out to investigate "d'ou
nait la necessite des institutions politiques" (p. 281 , my italics) . The
10. In Ouv cmpl, 3:279-34. This text has ben kown at least since
182 and was published for the frst time in 187. It is prt of C. E. Vaughan's still
authortative edition of Th Politial Wrting ofJ.J. Rus au (Cmbridge, Engnd,
1915) . The frst sect ion, entitled "Pemieres notions du corps social," is taken over in
the fnal version only from chapter 3 on (see p. 289). It deals with the development
leading up to the contract and thus establishes a transition beteen the Sond
Diurs and the Sl Cntrat.
.
PROMI SES ( SOCIAL CONTRCT)
259
section was subsequently omitted from the fnal versin, and this
omission has not made the interpretation of the Scial Contrat any
easier. It has made it difcul t, for example, to see in what way the
structure of the political entit established on a contractual basis
difers from that of an empirical or natural entity. The principle of
totalization that organizes the frst formal defnition of the contract
seems, in all respects, to be similar in kind to the organic link that
binds part to whole in a metaphorical synecdoche. The social pact is
determined as follows: "Everone puts his wll , his property, his
strength, and his person in common, under the direction of the
general will , and we receive as a body [en cors] each member as an
inalienable part of the whole" (p. 290) . The fnal version retains this
defnition, wth some changes in wording that do not detract from
its assumed holism. The metaphorical system that unites limb to
body, one to all , individual to group, seems frmly established. A few
paragraphs earlier in the same version, Russeau has described a
similar system: "There would exist a knd of common sensorium that
would control the correspondence of all the parts; public good and
evil would not only b the sum of particular virtues and vices as in a
simple juxtaposition [un simpl ag ation] , but it would reside in
the lin that unites them. It would therefore be larger than the sum,
and far from having public well-being derive from the happiness of
indivduals, it would b its souce" (p. 2) Y The distinction be
tween metonyic agregates and metaphorical totalities, based on
the presence, wthin the latter, of a "necessar link" that is lackng in
the former, is characteristic of all metaphorical systems, as is the
equation of this principle of totalization wth natural process. Afer
the deconstruction of the metaphorical model has taken place, the
attribute of naturalness shifs from the metaphorical totality to the
metonyic agregate, as was the case for the "state of nature" in the
Scon Dicourse or for "sensation" in the Pofesion d foi. However,
at this point in the argument , the evocation of this natural
synthesis-Rousseau makes the comparison with chemical com
pounds, whose properties are distinct from the properties of their
components-is not held up as the desirable wholeness that political
units must tr to emulate, but as the ver opposite: the are precisely
1 1 . The passge is crossed out in the manuscript (see 3: 1413 n. 4) for reasons
on which one can spculate. Its suppression i certainly not caused by a divergence
wth the general thrust of the argument, within which it fts prectly well .
260
ROUSSEAU
the misleading model afer which no sound political system should
be pattered. The syesthetic illusion of the common sensorium, just
as the concomitant illusions of a universal language and a golden
age, 1 2 is a myhical abrration of judgment devoid of truth and of
virtue. It becomes pernicious when it is used as the foundation of a
plitical societ: the entire polemic wth Diderot , in the manurit d
Gen, is directed toward the necessit of devising a model for a
political order that is not natural, in this sense ofthe term. Hence the
categorical reection, more explicit in the early version than in the
fnal text , of the family as a suitable political model, prcisely be
cause the family is based on natural ties. 13 In this respect , the family
is no better model for legalit than imperialistic conquest or the
enslavement of prisoners in time of war, and it is discussed under the
same rubric as these anarchic manifestations of power. The same is
true of the god-centered systems that occupy such a prominent place
in the Pofsin dfi and that, in this context, begin by being dis
missed altogether (although the will reappear in a diferent form
later on in the Sil Contrat) : the idea of a natural religion is as
absurd as the idea ofa natural law, and the text seems to b an even
sharper attack on the vicar of the Pofsion d foi than on Diderot :
" . . . if the notions ofa great Being and of natural law were innate in
all hearts, it would b quite superuous to instruct people in either of
them. . . . Let us therefore discard the sacred precepts of the vari
ous religions. Their abuse has caused as many crimes as their use can
prevent, and let us give back to the Philosopher a question that the
Theologian has treated only at the expense of mannd" (p. 286) . The
deconstruction of metaphorical totalities which, in "Du bonheur
public," starts out from the relationship between private and public
well-being here has a wder scope that encompasses all organic and
12. " . . . the happy existence of a golden age was always alien to the race of
man, either for not havng known it when i t was within reach or for having lost it
when the were capable of knowledge" (p. 283) .
13. Sil Contrat, frst version, p. 297. The fnal version (chap. 2, "D
premieres societes," p. 352) seems to reverse this when i t states: "The family is, if one
so wishes (si l'on vut ) , the frst model of political soieties." But the passage uses the
same metaphor of the "lin" negtively ("Stil l , the childn remain lined [li] to the
father for only as long as the need him in order to surive. As soon as this need
ceases , the natural link dissolves") and concludes that actual families are in fact not
natural but political institut ions: "If the remain together, it i no longer for natural
reasons but by their own wil l , and the family itself remains in existence only by
convention" ( p. 352) .
PROMI SES (SOCIAL CONTRCT) 261
theotropic ideologies. It is not carried out in a detailed analysis, as
was the case i some of the other tets, but assered sweepigy, as i
it could be taen for ganted. Ifthe formal defnition of the contract
then seems to relapse into the fgre which has just been decisively
condemned, then this can certainly not be true without frther qual
ifcation. The defnition is far from telling all there is to tell about the
structure ofthe contract , perhaps because a degree of complext has
been reached that no longer allows for defnitional language.
As the mechanics of the contractual convention are being elabo
rated in more detail , it becomes apparent that the constitutive power
of the contract , the manner in which it engenders entities, is no
dialectical sthesis or any other system of totaliztion. The general
will is by no means a synthesis of particular volitions. Rousseau starts
out instead from the opposite assertion and postulates the incom
patibility between collective and indivdual needs and interests, the
absence of any lins beteen the two sets offorces: the general will is
"a pure act of reason that operates without regard for the passions
[dn I sile d pasion] . . . " (p. 286) but "where is the person
who can thus sever himself from his own desires and . . . can he be
forced to consider the species in general in order to impse upon
himself duties of which he cannot perceive the lin with his ow,
particular constitution? (p. 286). It is clea that when Russeau, in
the next paragraph, speak of "the art of generalizing ideas" in order
to orient them toward the general will , the act of generalizing must
then have a ver diferent fgural structure than such metaphorical
processes as, for exmple, conceptualiztion, love, or even judgent.
The simplest way to gain insight into the process or function
that is being described may be by way of its most naive, spatial
version, in the section of the Scil Contrat entitled, in the mnu
crt d Gene, "Du domaine reel" (pp. 293-94) . Considered from a
geopolitical point of vew, the State is not primarily a set ofindivdu
als, but a specifc piece ofland; Russeau praises the wisdom implied
in the modern custom of calling a monarch Kng ofFane, or of
Engnd, rather than, as was the case in Antiquit, King of the
Persians or of the Macedonians. "By thus holding the land, [the
kngs] are certain to hold its inhabitants" (p. 293) . This terminolog
is said to be more precise bcause the orignal possesion of the land
is, in fact, arbitrar and "natual," in the anarchic sense ofthe term.
One could call it metonymic, simply based on the fact that one
happened to be in the proximit of this particular piece of terrain,
26 ROUSSEAU
and this "rigt offrst occupanc may be "less absurd, less repllent
than the right of conquest" but nevertheless, "on close eamination
. . . it is not more legitimate" (p. 293) .
A specifc piece oflanded proprt within the State i s the result
of a contractual convention that involves both the citizen and the
State; it is only when the State is thus involved that one can speak of
propert rather than of mere possession. Although propert, unlike
possession, will est and function within a legal context that is no
longer based on mere physical inequalit, it is not in itself more
legtmate, if only because the State itself, on which it depends, is
such an arbitrar entit: ". . . since the powers of the State are
incomparably larger than those of each indivdual , public possession
is, in practice, stronger and more irrevocable, without being more
legitimate" (p. 293) . The contractual instrument that is thus consti
tuted ests a a pradoxcal juxtapositon or iterference of relational
networks. On the one hand, as private property, objects of possession
used for the fulfllment of individual needs and desires voui ane] ,
the relationship between the owner and the land, or dwelling, is
entirely literal . It is perfectly defned in its identit by its objective
dimensions, and the inscribed signs by means of which these dimen
sions are designated (b it as a fence, or a a "nontrespassig sign) is
semantically unambiguous. A principle of functional identifcation
between the owing subject and the owed object is implied. This
identifcation, as we saw, is not natural and legtimate, but contrac
tual . There is nothing legitimate about propert, but the rhetoric of
propert confers the illusion of legitimac. The contract is self
refective; it is an ageement du mm au mm in which the land
defnes the ower and the ower defnes the land. One could say,
with equal justice, that the prvate owner contracts with himself or
that the private propert contracts with itself; the identit of the
ower is defned by the identit of the land. Thus it is that Marcel , in
Poust' s novel , understands the fascination of the prpe names of
the aristocrac because it is impossible to distinguish their names
from the geogaphical names of their landed estate. There can be no
more seductive form of onomastic identifcation. The fascination of
the moel is not so much that it feeds fantasies of material possession
(thoug it does this too, of course) but that it satisfes semiological
fantasies about the adequation of sign to meaning seductive enoug
to tolerate extreme forms of economic oppression.
PROMISES (SOCIAL CONTRACT) 26
On the other hand, Rousseau stresses that propert can also be
considered from a public point of vew as part of the rights and
duties of the State: "This is how the united and contiguous land of
private owers becomes public territor, and how rights of sover
eigt, extending from the subjects to the land they occupy, become
both 'real' and personal . . . " Cp. 293). But the relationship that
governs the public aspect of the proprt is not the same as the one
that determines its private identit. The diference becomes apparent
in the way in which the propertys "inside" relates to what lies be
yond its boundaries. When considered from the private point of vew ,
this relationship is still governed by patterns of genuine similarit.
The relation of one private propert to another is a relationship
between two units that are similarly constituted, and it therefore
sufces to respect the principle of this constitution in order to have a
reliable system of arbitration available in the adjudication of any
confict that may arise btween the two owners. Neighbors can have
any number of conficts with each other, but whether they want it or
not, the remain neigbors and not strangers; as far as their mutual
property rights are concerned, they can always be derived from the
proprietar status the have in common, in the form of a deed or any
other instrment of ownership. Local obscurities in the phrasing of
the deed can be clarifed and the deed is, in principle, a denomina
tive text that is entirely readable to all parties. Therefore, though it is
never guranteed, pace between neighbors is at least legally conceiv
able. But the same is not tre when the property is considered within
a context of public interests, especially when they involve the inter
ests of the State with regard to other States. The contractual constitu
tion of a State may or may not be similar to that of another, but this
question is irrelevant wth regard to territorial confict and integrit.
From that point of view, the other State is, per defnition, a hostile
stranger: ". . . the Greek ofen called peace treaties the treaties
established between two people who were not at war. The words for
stranger and for enemy have long been synonyms for many ancient
people, including the Rmans" Cp. 2 ). When considered privately,
propert is a structure based on similarity and on the integation of
shard needs and desires; when considered publicly, the same prop
erty functions as a structure of necessar estrangement and confict.
This hostility is the foundation of the State's political integrit and
can therefore be valorized positively: it protects propert "with all its
26
ROUSSEAU
strength against the outsider [contr l'Etran
g
er]" and it enables Rus
seau to speak approvingly of "what is admirable in this alienation"
(p. 29) .
From a rhetorical perspective, the interest of the structure is that
the same single entit (a specifc piece of land) can be considered as
the referent of two entirely divergent texts, the frst based on the
proper meaning engendered by a consistent conceptual system, the
second on the radical discontinuit and estrangement of noncom
parable systems of relationship that allow for no acts of judgment
and, consequently, for no stable meaning or identit. The semiologi
cal sstems at work within each of these systems are entirely difer
ent: the one is monological and controlled in all its articulations, the
second at the mercy of contingencies more arbitrar even than the
strength based on numerical power. Yet, in its absence, the frst
could never have come into being. The power of propert is vested
"in the distinction between the rights that the sovereign and the
ower have over the same fund" (p. 29). Behind the stabilit and
the decorm of private law lurk the "brigands" and the "pirates" (p.
28) whose acts shape the realities of politics between nations, the
most difcult adjustment being the necessit of considering these
mixed standards as entirely honorable (p. 288).
The pattern may seem crude and literal when it is applied to
material propert, but it perades all aspects of the political societ.
The social contract is best characterized, not by the conceptual lan
guage of its formal defnition, but by the dubl ra
pp
ort that we
found operative in the determination of propert rights and that also
characterizes the pact in its most fundamental form, of which prop
ert is only a derived, particular version. The expression dubl
r
pp
ort is used in a difcult and controversial passage that formu
lates a distinction in the degee to which the contract is binding for
the indivdual citizen as compared to the degree to which it is bind
ing for the sovereign: " ... the act of original confederation includes
a reciproal commitment btween the public and the private sector.
Each individual, contracting so to speak wth himself, is committed
in a double relationship, namely as member of the sovereig au
thorit with regard to individuals and as member of the State wth
regard to the sovereign authorit .. . " (p. 290). We know from
empirical experience that the individual is subjected to a more strin
gent legl contrl than the executive power which has much more
leeway in its actions and initiatives, in international politics, for
PROMISES (SOCIAL CONTRACT) 265
example, where it is expected to resort to war and to violence in a
manner that could not be tolerated in relationships between indi
viduals. Russeau accounts for this by stressing that the private inter
ests of the individual (which can be called his commitments toward
himself have nothing in common with his political, public interests
and obligations. The former do precisely not derive from the latter as
part derives from whole; the sentence "there is considerable difer
ence between a commitment toward oneself and a commitment
toward a whole of which one is a part" (p. 290) is true because no
metaphorical totalization is allowed to interene. Te double rela
tionship of the individual toward the State is thus based on the
coexistence of two distinct rhetorical models, the frst self-refective
or specular, the other estranged. But what the indivdual is estranged
from is precisely the executive activity of his ow State as souv
rin. This power is unlike him and foreign to him because it does
not have the same double and self-contradictor structe and
therefore dos not share in his problems and tensions. The souv
rain can consider himself "under one singe and identical relation
ship" and, with regard to any outsider, including the indivdual
citizen, it can become "a single Being or an individual" ("a l'egard de
l'etranger, il devent un Etre simple ou un indivdu" [po 291]). Unlike
the "individual," who is always divided within himself, the executive
is truly in-dividual, un-divided.
The passage becomes clearer in its implications if one tas into
account the precise meaning of the terminolog. Wat Russeau calls
the "souverain" ( which can, with some historical hindsight, be trans
lated as the executive power) is, of course, not a person but speci
fcally the political body when it is active as distinct from this same
body as a mere entity, the carrier, or ground, as it were, of the action
that it makes possible by its existence: " . . . the sovereign authority,
by its ver nature, is only a moral person, whose exstence can only
be abstract and collective. The idea that is linked to this term cannot
be equated with that of a single individual (pp. 29-95). ''This public
person thus constituted by the union of all the others . . . is called by
its members the Stat [Ett] when it is passive, the Sverig [Suv
rin] when it is active . . . " (p. 290). The indivdual's private will
(like his private property) is clear and comprehensible in itself, but
devoid of any general interest or signifcation beond himself. The
same is still true of the relationship beteen several private volitions.
Only with the "double relationship" does the possibilit of generali-
26
ROUSSEAU
ztion come into being but, at the same moment, the continuit
btween purpose and action that remained preseIed on the private
level is disrpted. It follow that the divergence which prevails,
within the State, in the relationship between the citizen and the
executive is in fact an unavoidable estrangement between political
rigts and laws on the one hand, and political action and histor on
the other. The gounds for this alienation are best understood in
terms of the rhetorical strcture that separates the one domain from
the other. The passage, as is well kown, was to be one of the main
considerations in condemning Russeau as politically subvrsiv. 14 It
proves however that he had a much more developed sense of political
praxis than the magistrates of the cit of Gneva; he certainly never
meant to imply that the executiv has the rght to change the con
stitution at its own will, but merely to caution against the fact that it
would always be tempted and have the power to do just that, and
that therefore the State needs legislation to protect it against the
prsistent threat of its own executive branch. And, as his admiration
for Moses indicates, Rousseau is equally convinced of the need for
durable institutions or States. But precisely because it is not rooted in
the contract itself, durabilit has to b legislated. The &xial Contrat
does not warrant belief in a suprahistorical political model that, in
the words of the 1738 Eit d Meiation of the State of Gneva,
would make the political State "peretual."15 For this would reduce
the double structure of the constitutional text to a monological sig
nifcation and cause the State to relapse into the kind of aberrant
natural model of which the end of the Scon Dicours gives a
fctional description. The declaration of the "permanence" of the
State would thus greatly hasten its dissolution. It follows however
that the meaning of the contractual text has to remain suspended
and undecidable: "there can be no fundamental Law that is binding
for the entire boy of the people" (p. 291) and". . . since the deci
sions of the executive, as executive, concern only itself, it is alway
free to change them" (p. 316). Revolution and legalit by no means
cancel each other out, since the text of the law is, per defnition, in a
14. Se Drathts note (3:147) on the objections of Jean-Rbert Tronchin,
Chief Prosecutor for the Rpublic of Gneva and hi complaint that, for Russau,
"the constitutive laws of all governments were never to be considered a irable."
This led to the condemnation of the Sl Cntrat as "destrctive of all gver
ment."
15. Se]. S. Spin,Jean Jaqu Rus au et Gn (Paris, 1934). p. 23.
PROMISES (SOCIAL CONTRACT) 267
condition of unpredictable change. Its mode of existence is necessr
ily tempral and historcal, thoug in a strictly nonteleologcl sns.
The strcture of the entit with which we are concerned (be it
as propert, as national State or any other political institution) is
most clearly revealed when it is considered as the general form that
subsumes all these particular versions, namely as legal t. The frst
characteristic of such a text is its generalit: " .. . the object of the
Law must be general, like the wll that dictates it. It is this double
universalit that determines the legalit of the Lw." At frst sight, it
seems that this generalit is rooted in the selective applicabilit of the
law to the part that makes up the political whole, as the exclusion of
the part that does not partae of this whole: "This is why the general
wll of an entire people is not general for a foreign individual [un
pariulier euanger], for this individual is not a member of this
people" (p. 327). But it turns out that the etrangement is not (as it
still misleadingly appeared to b in the case of territorial propert
models) the result of some spatial, temporal, or psychologcal
nonpresence, but that it is implied b the ver notion of particularit
itself. To the extent that he is particular, any individual is, as indi
vidual, alienated from a law that, on the other hand, exsts only in
relation to his individual being. "For at the moment that a pople
consider a particular object, een i it b on of it ow members, a
relation is created between whole and part that makes them into two
separate beings. The part is one of these being, the whole minus this
part the other. But the whole minus a part is not the whole and as
long as the relationhip persists, there is no whole, only to unequal
parts" (p. 327, my italics). This statement is repeated whenever the
mode of applicabilit of the law to particular citizens is under dis
cussion. ". . . there is no general will acting upon a particular ob
ject" but the particularit of the legl subject, to which the law is
made to apply, is independent of its being inside or outside the
precinct of the State; the categories of inside and outside do not
function as the determining principle of an unavoidable estrange
ment. That the era-muros individual is estranged from the law is
obvious: "If [the particular object] is outside the State, a wll that is
estranged from him is not general in rlation to him ... " (p. 378).
But the same applies necessrily to the individual intr-mur
,
sim
ply because he is individual or particular: "Indeed, this particular
object is either within the State, or wthout it. . .. If the object is
within the State, then it is a part of it. A relation is then created
26
ROUSSEAU
between the whole and the part that makes them into two separate
beings. The part is one of them, the whole minus the part the other.
But the whole minus a part is not the whole . . . etc. " (pp. 378-79).
From the point of view of the legal text, it is this generalit
which ruthlessly rejects any particularization, which allows for the
possibilit of its coming into being. Within the textual model, par
ticularization coresponds to reference, since reference is the applica
tion of an undetermined, general potential for meaning to a specifc
unit. The indiference of the text wth regard to its referential mean
ing is what allows the legl text to proliferate, exactly as the preor
dained, coded repetition of a specifc gesture or set of gestures allows
Helen to weave the stor of the war into the epic. As a text, the Sci
Conta is unusual among Russeau's work because of its imper
sonal, machinelike systematicit: it takes a few ke terms, programs
a relationship between them, and lets mere syntax take its course. It
is, for instance, the only Russeau text to make explicit and repeated
use of mathematical ratios. B suppressing the genealog of the key
terms from the fnal version, this quasi-mechanical pattern becomes
even more evident. "I shall go directly to my subject' without frst
demonstrating its importance" Russeau announces at the start of
the fnal text (p. 351), but the early version still felt the need to make
explicit what is taken for ganted later on: "I descrbe the main
springs [of the social body] and its various parts, and I assemble them
into place. I make the machine ready to go to work; others, wiser than
I am, will regulate its movements" (p. 281) .
We have moved closer and closer to the "defnition" of t, the
entit we are trng to circumscribe, a law bing, in its facticit,
more lie an actual text than a piece of propert or a State. The
system of relationships that generates the text and that functions
independently of its referential meaning is its gammar. To the ex
tent that a text is gammatical, it is a logcal code or a machine. And
there can be no agammatical texts, as the most grammatical of
poets, Mallarme, was the frst to ackowledge. 16 Any nongrammati
cal text wll always be read as a devation from an assumed gram
matical norm. But just as no text is conceivable wthout gammar,
no grammar is conceivable without the suspension of referential
16. "Quel pivot, j'entends, dns ces contrastes, a l'intelligibilite? il faut un
garantie-La Sytaxe-" (L Myte dn 1 lt, Pieiade edition [Paris, 195], p.
385).
PROMISES ( SOCIAL CONTRCT) 269
meaning. Just as no law can ever be witten unless one suspends any
consideration of applicabilit to a particular entit including, of
course, oneself, grammatical logic can fnction only if its referential
consequences are disregarded.
On the other hand, no law is a law unless it also applies to
particular individuals. It cannot be lef hangng in the air, i the
abstraction of its generalit. Only by thus referring it back to particu
lar pras can the jutie of the law be tested, exactly as the jutse of
any statement can only be tested b referential verifability, or by
deviation from this verifcation. For how is justice to b determined if
not by particular reference? "Wy is the general will always rght,
and why do all citizens constantly desire the well-being of each, if it
were not for the fact that no one exists who does not secretly appro
priate the term each and think of himself when he votes for all [il ny
a peronne qui ne s'apprpr en secret ce mot chacun et qui n songe
a lui-mm en votnt pour tou]? Which proves that the equalit of
right and the notion of justice that follows from it derive from the
preference that each man gives to hi
m
self, and therefore from the
nature of man" (p. 306).17 There can be no text without grammar:
the logic of gammar generates texts only in the absence of referen
tial meaning, but ever text generates a referent that subverts the
grammatical principle to which it owed its constitution. What re
mains hidden in the everday use of language, the fundamental
incompatibility between grammar and meaning, becomes explicit
w
h
en the linguistic structures are stated, as is the case here, in
political terms. The preceding passage makes clear that the incom
patibilit between the elaboration of the law and its application (or
justice) can only be bridged by an act of deceit. "S'approprier en
secret ce mot chun" is to steal from the text the ver meaning to
which, according to this text, we are not entitled, the paticular I
which destros its generalit; hence the deceitful, covert gesture "en
secret," in the foolish hope that the thef will go unnoticed. Justice is
unjust; no wonder that the language of justice is also the language of
gilt and that, as we know from the Coneion, we never lie as
much as when we want to do full justice to ourselves, especialy in
17. Russeau suppressd the all-important specifcation "en secret" in "s'ap
proprier [en secret] ce mot chun . . . " in the fnal version (p. 373). The self
censorship that operates between earlier and later versions of texts cn, in cases such
as this, reveal more than it conceals: how could one b more secretive than b tlng
to hide "the secrt"?
27
ROUSSEAU
self-accusation. The divergence between grammar and referential
meaning is what we call the fgal dimension of langage. This
dimension accounts for the fact that two enunciations that are
lexicologically and gammatically identical (the one being, so to
speak, the quotation of the other and vice versa) can, regardess of
context, have two entirely divergent meaning. In exactly the same
way Rousseau defnes the State or the law as a "double relationship"
that, at closer examination, turns out to be as self-destrctive as it is
unavoidable. In the description of the structure of political societ,
the "defnition" of a text as the contradictor interference of the
grammatical with the fgral feld emerges in its most systematic
form. This is not unexpected, since the political model is necessarily
diaphoric and cannot pretend to ignore the referential moment en
tirely. We call t any entit that can b considered from such a
double perspective: as a generative, open-ended, non-referential
grammatical system and as a fgral system closed of by a transcen
dental signifcation that subverts the grammatical code to which the
tet owes its exstence. The "defnition" of the text also states the
impossibilit of its exstence and prefgres the allegorical narratives
of this impossibilit.
In the Scil Contat
,
the model for the structural description of
textualit derives from the incompatibilit beteen the formulation
and the application of the law, reiterating the estrangement that
exists between the sovereign as an active, and the State as a static,
principle. The distinction, which is not a polarit, can therefore also
be phrased in terms of the diference beteen political action and
plitical prescription. The tension beteen fgral and grammatical
language is duplicated in the diferentiation bteen the State as a
defned entit (Etat) and the State as a principle of action (Souverain)
or, in lingistic terms, beteen the constative and the perormative
function of language. A text is dened by the necessit of considering
a statement, at the same time, as performative and constative, and
the logical tension between fge and grammar is repeated in the
impossibilit of distinguishing beteen two linguistic functions
which are not necessarily compatible. It seems that as soon as a text
kows what it states, it can only act deceptively, like the thievng
lawmaer in the Sl Contrat
,
and if a text does not act, it cannot
state what it knows. The distinction beteen a text as narrative and
a text as theor also belong to this feld of tension.
Especially in the fnal version, with the conceptual genealog
PROMISES (SOCIAL CONTRACT) 271
and the metaphorical deconstruction omitted, the Sial Contrat
does not seem to be a narrative but a theor, a constitutional
machine to which Rousseau was to resort for the elaboration of
specifc constitutions. If this were the case, then the text of the law
and the law of the text would fully coincide and generate both the
Sial Contrat, as a master text, and a set of contractual rules on
which the constitution of any State could be founded, or from which
the suitabilit of a given territor to be made into a State could be
deduced. It turns out, however, that the "law of the text" is too
devous to allow for such a simple relationship between model and
example, and the theor of politics inevitably turns into the histor,
the allegor of its inability to achieve the status of a science. The
passage from constative theor to performative histor is ver clearly
in evidence in the Scl Contrat. The text can b considered as the
theoretical description of the State, considered as a contractual and
legal model, but also as the disintegration of this sme model as soon
as it is put in motion. And since the contract is both statutor and
operative, it will have to be considered from this double perspective.
The legl machine, it turns out, never works exactly as it was
programmed to do. It always produces a little more or a little less
than the original, theoretical input. Wen it produces more, things
go almost too well for the State: " . . . the more [the] natural forces
[of man] are withered away and eliminated, and the more his ac
quired forces are stable and powerful, the more solid and perfect the
institution will be. Legislation reaches the highest point it can reach
when each citizen can act only through the others and when th
aquired power of th whol i eual or superior to the sum of the
natural forces of all indivduals" (p. 313, my italics). The result of
this supplementar efcienc of the political process is stated in
(metaphorical) terms that are not entirely reassuring, neither physi
cally nor epistemologically, since the suggest a substitutive process
that is far from harmless: to found a State is "to substitute a frag
mentar and moral existence for a physical and autonomous one,"
and this reductive substitution is called to kill, to annihilate, and "to
mutilate, so to speak, the human constitution in order to strengthen
it" ("mutiler en quelque sorte la constitution de l'homme pour la
renforcer" [po 313]).18 A somewhat crptic and isolated note by Rous
seau would seem to ft the situation: "I created a people, and 1 was
18. Te fnal version deletes "mutiIer" and replces it by the innocuous "al
hrer." Te emendation is pointed out, wthout comment, by Derathe (p. 146, n. 7).
272 ROUSSEAU
unable to create men" ('j'ai fait un peuple et n'ai pu faire des hom
mes" [po 500]). The thing to worr about is perhaps not so much the
redoubtable power that the State can generate as the fact that this
power is not necessarily equal to the power that went into its produc
tion. For at other moments, the supplementar or diferential struc
ture of the input/output relationship can take on a negative as well as
a positive sign: "The general will is rarely that of all, and the public
strength is always less than the sum of individual strengths. As a
result, we fnd in the wheels of the State an equivalent of the princi
ple of inertia in machines [on trouv dn les rsor d l'etat un
equivalnt au fottemnt ds mhins]. This counterforce has to be
kept to a minimum or, at the ver least, it must be computed and
deduced beforehand from the total power in order to set up a proper
ratio between the means one uses and the efect one wshes to pro
duce" (p. 287).19 The transformation of the generative power of
theor and of grammar into a quantitative economy of ls,
2
0 a kind
of political thermodyamics governed by a debilitating entropy, il
lustrates the practical consequences of a linguistic structure in which
gammar and fgre, statement and speech act do not converge.
Rgardless of whether the diferentiation engenders excess or
default, it always results in an increasing devation of the law of the
State from the state of the law, between constitutional prescription
and political action. As in the Prfsion d fi, this diferential struc
ture engenders an afectivit and a valorization, but since the difer
ence is one of epistemological divergence between a statement and
its meaning, the afect can never be a reliable criterion of political
value judgment. A we already kow from "Du bonheur public," the
eudaemony of politics is not an exact science: "In order for everone
to want to do what he has to do according to the commitment of the
social contract, everone has to kow what he should want. What he
should want is the common good; what he should avoid is public
evl. But since the existence of the State is only ideal and conven
tional, its members possess no common and natual sensibilit
19. "One cannot avoid, in politics as in mechanics, acting weaky or slowly,
and losing strength or time" (&il Contra, frst version, p. 29).
20. See also Er sur l'Ab d Sint-Pr: "Consider to what extent, in the
a
gegate of the body politic, the public power is inferior to the sum of particulr
powers, how much inertia there is, s to spak, in the play of the entire machine, and
one will have to conclude that the weakest man disposes proportionately of more
strength for his surval than the strongest State for his" (3:60).
PROMISES (SOCIAL CONTRCT) 2'
through which, wthout mediation, they are forewarned by a
pleasurable impression of what is useful to the State and b a painful
impression of what could harm it" ("ses membres n'ont aucune
sensibilite natuelle et commune, pa laquelle, immediatement aver
tis, ils reloivent une impression ageable de ce qui lui est utile, et une
impression douloureuse sitot qu'il est ofense" [po 309]. The af
fective coe is unreadable, which is equivalent to stating that it is
not, or not merely, a code.
The discrepanc within the contractual model (here claimed to
be the lingistic model in general) will necessarily manifest itself
phenomenologically, since it is defned, in part, as the passage, how
ever unreliable, from "pure" theor to an empirical phenomenon.
The noncoincidence of the theoretical statement wth its phenom
enal manifestation implies that the mode of existence of the con
tract is temporal, or that time is the phenomenal categor produced
by the discrepanc. Considered performa tively, the speech act of the
contractual text never refers to a situation that exists in the present,
but signals toward a hyothetical futue: "Far from preventing the
evils that attack the State, [the members of the State] are rarely on
time to remedy them when the begin to perceive their efects. One
has to foresee them well in advance in order to avoid or to cure
them" Cp. 309). Al laws are future-oriented and prospective; their
illocutionar mode is that of the prmie.
2
1
On the other hand, ever
promise assumes a date at which the promise is made and without
which it would have no validit; laws are promissor notes in which
the present of the promise is always a past with regard to its realiza
tion: ". . . the law of today should not be an act of yesterdays
general wll but of todays; we have not committed ourselves to do
what the people wanted but what the want. It follows that when
the Law speaks in the name of the people, it is in the name of the
people of today and not of the past" (p. 316). The defnition of this
"people of today is impossible, however, for the eternal present of
the contract can never apply as such to any particular present.
The situation is wthout solution. In the absence of an etat
preent, the general wll is quite literally voiceless. The people are a
21. In Th Geneal of Morl, Nietzsche also derives the notion of a tran
scendental referent (and the specifcit of "man") from the possibility of makng
promises. Se, for example, Friedrich Nietzsche, Werk, edited by Krl Shechta
(Munich, 1955), 2:805, 826.
274 ROUSSEAU
helpless and "mutilated" gant, a distant and weakened echo of the
Polyhemos we fst encountered in the Son Diours.
22
"Does the
body political possess an orgn with which it can state [nr] the
will of the people? Wo will gve it the necessr foresight to shape
the people's actions and to announce them in advance, and how will
it pronounce them when the need arises? How could a blind mob,
which ofen does not kow what it wants because it rarely kows its
own good, can out by itself as huge and difcult an enterrse as
the promulgation of a system of Law" Cp. 380)? Yet it is this blind
and mute monster which has to articulate the promise that will
restore its voice and its sigt: "The people subject to the Lw must be
the authors of the Law" Cp. 390). Only a subterfuge can put this
paralysis in motion. Since the system itself had to be based on deceit,
the mainspring of its movement has to be deceitful as well.
The impostor is clearly enough identifed: Rosseau calls him the
"lawgiver." It has to be an individual, since only an indivdual can
have the sigt and the voice that the people lack. But this individual
is also a rhetorical fgre, for his abilit to promise depends on the
metaleptic reversal of cause and efect: "For a people to appreciate
the sound maxms of politics and to follow the fundamental rles of
political reason [l raion d'Etat], efect should become cause, and
the social spirit that the institutions are to produce should preside
over their elaboration. Men should be, prior to the laws, what they
are to become through them" Cp. 383). The metaphor engendered b
this metalepsis is equally predictable. It can only be God, since the
temporal and causal reversal that puts the realization of the promise
before its utterance can only occu within a teleological system
oriented toward the convergence of fgure and meaning. Since the
Sil Contact is nothing of the sort, it is entirely consistent that it
should introduce the notion of divne authorit at this particular
point and have to defne it as a simulacrm: "If prideful philosophy
and blind partisan spirit continue to regard [the lawgiver] as a fortu
nate impostor, tre political minds admire in the institutions they
created, the forceful genius that presides over enduing laws" Cp.
38) . When the trly political mind is also a philosopher, he will no
longer b "pridef," but this lawgvr will b no less of an impostor,
albeit no longer a fortunate one. The metaphorical substitution of
one's own for the divine voice is blasphemous, although the necessit
22. S Chapter 7, p. 153, n. 29.
PROMISES (SOCIAL CONTRCT) 275
for this deceit is as implacable as its eventual denunciation, in the
future undoing of any State or any political institution.
Is Russeau himself the "lawgiver" of the Sil Contrt and his
treatise the Deuteronomy of the modem State? If this were the case,
then the Sl Contrat would become a monological referntial
statement. It could not be called an allegor, nor a text in our sense,
since the exposure of the deceit would have to come from outside
evidence not provded by the text itself. Since it implicitly and
explicitly denies, in chapter 7 of book 2 ("D legislateur") and again
in the related chapter 8 of book 4 ("De la religion civile"), any form of
divne inspiration for itself, it is clear that Russeau does not identif
himself with any of the major legslators, be it Moses, Lycurgus, or
Christ; instead, by raising the suspicion that the Sermon on the
Mount may be the Machiavellian invention of a master politician, he
clearly undermines the authority of his ow legislative discourse.
Would we then have to conclude that the Sil Contrat is a decon
structive narrative like the Son Dicours? But this is not the case
either, because the Sl Contrat is clearly productive and genera
tive as well as deconstrctive in a manner that the Scon Dicour is
not. To the extent that it never ceases to advocate the necessity for
political legslation and to elaborate the principles on which such a
legslation could be based, it resorts to the principles of authority
that it undermines. We kow this structure to b characteristic of
what we have called allegories of unreadability. Such an allegor is
metafgural: it is an allegor of a fgur (for example, metaphor)
which relapses into the fge it deconstructs. The Sil Contrt
falls under this heading to the etent that it is indeed structued like
an aporia: it persists in performing what it has show to be impossi
ble to do. A such, we can call it an allegor. But is it the allegor of a
fgure? The question can be answered b asking what it is the Sil
Contrat performs, what it keeps doing despite the fact that it has
established that it could not b done. What the Pofsion dfi keeps
doing is to assert the metaphorical analog between mind and na
ture against which the text has generated its own argment; it keeps
listening, in other words, to the voice of conscience (or of G)
afectively, althoug it no longer can believe it. Wat Julie keeps
doing, at the end of part 3 of the novel, is to "love" Saint-Pux and
G as if they were interchangeable. To listen and to love are referen
tial transitive acts that are not self-positing. What the Sil Contrat
keeps doing however is to promise, that is, to perorm the ver
27
ROUSSEAU
illoutionar speech act which it has discredited and t o perform i t in
all its textual ambiguit, as a statement of which the constative and
the performative functions cannot b distinguished or reconciled.
That the Scil Contat denies the right to promise is clear
from the fact that the legislator has to invent a transcendental prin
ciple of signifcation called G in order to perform the metalepsis
that reverses the temporal pattern of all promissor and legal state
ments. Since Gd is sid to be, within this perspective, a subterfuge,
it follows that the Scil Contat has lost the right to promise any
thing. Yet it promises a great deal. For example: " ... far from
thinkng that we can have neither vrtue nor happiness and that
providence [l cil] has abandoned us without shelter to the degrad
tion of the species, let us extract from evl itself the remedy that must
cure it. Let us if possible improve by new institutions [d nouvll
ascition] the shortcomings of society in general. . . . Let our
interlocutor see in a better constitution of things the reward for
virtuous deeds, the punishment of evil and the harmonious accord of
justice and well-being" (p. 28) . Or: " ... it becomes obvious that
individuals do not really give up anything when they enter into the
social contract. Their new situation is genuinely preferable to the old
one, before the contract. Instead of an alienation, the have ex
changed an uncertain and precarious way of life against a better and
more secure one; instead of natural independence, the now have
freedom; instead of the power to harm others, the now have their
ow securit; and instead of thei indivdual strength, which others
might overcome, they now have rights which the soial union makes
invincible" (p. 375). Sveral other instances could be quoted, some
explicit, some all the more sugestive because the are all-pervasive
in their connotations; it is impossible to read the Scil Cntat
without expriencing the exhilarating feeling inspired by a frm
promise.
The reintroduction of the promise, despite the fact that its im
possibilit has been established (the pattern that identifes the Scil
Contat as a teual allegor), does not occur at the discretion of the
witer. We are not merely pointing out an inconsistenc, a weakess
in the text of the Scil Contrt that could have ben avoided by
simply omitting sentimental or demagogical passages. The point is
not that the Scil Contrat relapses into textual activism because it
does so explicitly, in sections and passages that can be isolated and
PROMISES (SOCIAL CONTRCT) 27
quoted by themselves. Even without these passges, the Sl Cn
trat would still promise by inference, perhaps more efectively tha
if Rusau had not had the naivete, or the good faith, to promis
openly. The redoubtable efcac of the text is due to the rhetorcal
moel of which it is a version. This moel is a fact of langage over
which Rousseau himself has no control. Just as ay other reader, he is
bound to misread his text as a promise of political change. The eror
is not within the reader; langage itself dissociates the cogition
from the act. Di Sprah vrspriht (sih); to the extent that is neces
srily misleading, langage jut as necessarly conves the promise of
its ow truth. This is also why textual allegories on this level of
rhetorical complexity generate histor.
12
Excuses
(Crein)
POLITICAL AND AUTOBIOGRAPHICAL TEXTS HAVE IN COM
mon that they share a referential reading-moment exlicitly built in
wthin the spectrum of their signifcations, no matter how deluded
this moment may be in its mode as well as in its thematic content:
the deadly "horn of the bull" referred to by Michel Leirs in a te that
is indeed as political as it is autobiogaphical.1 But whereas the
relationship between cognition and performance is relatively easy to
grasp in the case of a temporal speech act such as promie-which,
in Rousseau's work, is the model for the Sl Conta-it is more
complex in the confessional moe of his autobiographies. By reading
a central passage from the Confe in, I attempt to clarit the rela
tionship between critical procedures that stat out from the dis
course of the subject and procedures that start out from political
statements.
Among the various more or less shamefl and embaassing
scenes from childhood and adolescence related in the frst three
book of the Confe in, Rousseau singed out the episode of Marion
and the ribbon as of particular afective sigifcance, a truly primal
scene of lie and deception strategcaly placed in the narrative and
told wth special panh. We are invited to believe that the episoe
was never revealed to anyone prior to the privleged reader of the
Confesin "and ... that the desie to free myself, so to spea, from
this weight has geatly contributed to my resolve to wite my confes
sions" (86].
2
Wen Russeau retuns to the Confsin i the later
Fourh R, he agin singes out this same episoe as a paradig-
1. "D la litterature consideree comme We tauomachie," in Michel Li,
L'ae d'mme (Paris: Gallimard, 19). The essy dates from 195, immedatel
afer the war.
2. Page nubr a from J. J. Russau, Ouv cmpl, L cnin,
aut te autbiphu, 00. Brr Gagebin and Marcel Rymond (Par:
Gallimard [Bibliotheque de la Pleaide], 1959), vol. 1. The pssg concludes Bok II
of the Conein and appars on pp. 85-87.
278
EXCUSES (CONFESSIONS) 27
matic event, the core of his autobiographical narrative. The selection
is, in itself, as arbitrar as it is suspicious, but it provdes us with a
textual event of undeniable exegetic interest: the juxtaposition of two
confessional texts linked together by an explicit repetition, the con
fession, as it were, of a confession.
The episode itself is one in a series of stories of pett larceny, but
with an added twist. While employed as a servant in an aristoratic
Tin household, Russeau has stolen a "pink and silver colored
ribbon." Wen the thef is discovered, he accuses a young maidser
vant of havng given him the ribbon, the implication being that she
was trng to seduce him. In public confrontation, he obstinately
clings to his stor, thus casting irreparable doubt on the honest and
the moralit of an innocent girl who has never done him the slightest
bit of harm and whose sublime good nature does not even finch in
the face of dastardly accusation: "Ah Russeau! I took you to be a
man of good character. You are making me ver unhappy but I
would hate to change places with you" (85). The stor ends badly,
wth both characters being dismissed, thus allowing Russeau to
speCUlate at length, and with some relish, on the dreadful things that
are bound to have happened in the subsequent career of the hapless
girl.
The frst thing established by this editng narrative is that the
Confe in are not primarily a confessional text. To confess is to
overcome guilt and shame in the name of truth: it is an epistemolog
ica
i
use of language in which ethical values of good and evil are
superseded by values of trth and falsehood, one of the implications
being that vces such as concupiscence, env, greed, and the like are
vices primarily because they compel one to lie. By stating things as
the are, the economy of ethical balance is restored and redemption
can start in the clarfed atmosphere of a trth that dos not hesitate
to reveal the crime in all its horror. In this case, Russeau even adds
to the horror by conjuring up, in the narrative of the Cnfsin as
well as that of the Pma, the dire consequences that his action
may have had for the victim. Confessions occur in the name of an
absolute truth which is said to exst "for itself ("pour elle seule,"
[1028]) and of which particular truths are only derivative and secon
dar aspects.
But even within the frst narrative, in Book II of the Con
fsion, Rousseau cannot limit himself to the mere statement
of what "really happened, although he is proud to daw attention
2BO
ROUSSEAU
to the fllness of a self-accusation whose candor we are never sup
posed to suspect: "I have been ver thorough in the confession 1 have
made, and it could certainly never be said that I tried to conceal the
blackess of my crime" (86). But it does not sufce to tell all. It is not
enough to cnfe, one also has to ecus: "But I would not fulfll the
purpose of this book if I did not reveal my inner sentiments as well,
and if I did not fear to ecuse myself by means of what conforms to
the trth" ("que je [ne] craignisse de m'ecuer en ce qui est con
forme i la verite" [86, my italics]). This also happens, it should be
noted, in the name of trth and, at frst sight, there should be no
confict between confession and excuse. Yet the language reveals the
tension in the expression: crinre de m'excuser. The only thing one
has to fea from the excuse is that it will indeed exculpate the confes
sor, thus making the confession (and the confessional text) redun
dant as it originates. Qui s'acuse s'ecuse; this sounds convncing
and convenient enoug, but, in terms of absolute truth, it ruins the
seriousness of any confessional discourse by mang it self-destruc
tive. Since confession is not a reparation in the realm of practical
justice but exists only as a verbal utterance, how then are we to
kow that we are indeed dealing with a tru confession, since the
recognition of guilt implies its exoneration in the name of the same
transcendental principle of trth that allowed for the certitude of
gilt in the frst place?
In fact, a far-reaching moifcation of the orgnizing principle
of truth occurs between the two sections of the narrative. The truth
in whose name the excuse h to be stated, even at Rousseau's as
sumed "corps defen
d
ant," is not structured like the truth principle
that governs the confession. It does not unveil a state of being but
states a suspicion, a possible discrepanc that might lead to an im
possibilit to know. The discrepanc, of course, is between the "sen
timent interieur" that accompanied (or prompted?) the act and the
act itself. But the spatial inside/outside metaphor is misleading, for it
articulates a diferentiation that is not spatial at all. The distinction
between the confession stated in the mode of revealed truth and the
confession stated in the mode of excuse is that the evdence for the
former is referential (the ribbon), whereas the evidence for the latter
can only be verbal. Rousseau can convey his "inner feeling" to us only
if we take, as we say, his word for it, whereas the evdence for his
thef is, at least in theor, literally available.
3
Whether we believe him
3. This is so evn within the immediate situation, when no actual text is
EXCUSES (CONFESSIONS)
21
or not is not the point; it is the verbal or nonverbal natue of the
evidence that maes the diference, not the sincerit of the speaker or
the gllibilit of the listener. The distinction is that the latter process
necessarily includes a moment of understanding that cannot b
equated with a perception, and that the logic that governs this mo
ment is not the sme as that which governs a referential verifcation.
Wat Rousseau is saying then, when he insists on "sentiment inter
ieur ," is that confessional langage can b considered under a double
epistemologcal perspective: it functions as a verifable referential
cognition, but it also functions as a statement whose reliabilit can
not be verifed by empirical means. The convergence of the two
modes is not a priori given, and it is bcause of the possibilit of a
discrepanc between them that the possibilit of excuse arises. The
excuse articulates the discrepanc and, in so doing, it actually asserts
it as fact (whereas it is only a suspicion). It believes, or pretends to
believe, that the act of stealing the ribbon is both this act as a
physical fact (he removed it from the place where it was and put it in
his pocket, or wherever he kept it), as well as a certain "inner feeling
that was somehow (and this "how" remains open) connected with it.
Moreover, it believes that the fact and the feeling are not the same.
Thus to complicate a fact certainly is: to act. The diference beteen
the verbal excuse and the referential crime is not a simple opposition
between an action and a mere utterance about an action. To steal is
to act and includes no necessar verba elements. To confess is dis
c
u
rsive, but the discourse is governed by a principle of referential
verifcation that includes an extraverbl moment: even if we confess
that we said something (as opposed to did), the verfcation of this
verbal event, the decision about the truth or falsehood of its occu
rence, is not verbal but factual, the kowledge that the utterance
actually took place. No such possibilit of verifcation exsts for the
excuse, which is verbal in its utterance, in its efect and in its author
it: its pupose is not to state but to convnce, itself an "inner" process
to which only words can bar wtness. A is well known at least since
Austin,
4
excuses are a complex instance of what he termed perform-
present. Someone's sentiments a accesible only through the medium of mimicr,
of geturs that ruire deciphering and function a a lngage. That th deciphering
is not necessrly reliable is clear from the fact that the facial expression of, say, a
thief at the moment he is caught rd-handed is not likely to weigh heavily as
evdence in a court of law. Our ow sentiments are available to us only in the same
manner.
4. Se, for example, J. L. Austin, "Prformative Utterances" and "A Plea for
ROUSSEAU
ative utterances, a variet of speech act . The interest of Rousseau's
tet is that it exlicitly functions prformatively as well as cogni
tively, and thus gives indications about the structure ofperformative
rhetoric; this is already established in this text when the confession
fails to close of a discourse which feels compelled to modulate from
the confessional into the apologetic mode. 5
Neither does the performance of the excuse allow for a closing
of of the apologetic text, despite Rousseau's plea at the end of Book
II: "This is what I had to say on this matter. May I be allowed never
to mention it again" (87). Yet , some ten years later, in the Fourh
Rer, he tells the entire stor all over again, in the context of a
meditation that has to do wth the possible "excusabilit of lies.
Clearly, the apolog has not succeeded in becalming his ow guilt to
the point where he would be allowed to forget it. It doesn't matter
much, for our purpose, whether the guilt truly relates to this particu
lar act or if the act is merely mde to substitute for another, worse
crime or humiliation. It may stand for a whole series of crimes, a
general mood of guilt, yet the repetition is signifcant by itself what-
Excuses," in Phlosophial Papes, ed. J. o. Urmson and G. J. Warnok (Oxford,
1961) .
5. The usual way of dealing with this recurent patter in Russau's wrtin@
is by stressing the bad faith of his commitment to a rral d l'intntn, the ethical
stance for which he was taken severely to task by Sartre. In hi commentar on the
pa g, Mel Rymond, thoug less sre, tas the sme approach: "B rev
ing his 'inner feeling' ['dpoitns intru'] which were good . . . it appeas
that afer havng stigatize hi misdeed he gradually begns to justif i t. The same
giding and swerng motion can be obsered more than once in the Confsins,
especially when Russeau accounts for the abndonment of hi children. He is
alway led to distinguish the intent from the act" ( 1273-74) . It can, however, b
shown that Russeau's ethics is much rather a moml d pmtiu than a rral d
l'intntion, and that this analysis therefore dos not account for the genuinely pre
Kntian interest of his ethical language and theor. The extensive possibilities of bad
faith engendered by the distinction beteen the actual event and the inner feeling
a abundanelt prsnt thugout Rusau, but the don't gor the mor pu
zling ad interting movements and coinages of the te. Wether the link bten
"inner" feeling and "outer" action can be called intentional is precisely the buren of
the interretation and cannot b assrted without frther evdence. Ifwe are rigt in
syg that "qi s'a s'e," then the rltionship btween confession and
excus is rhetorical prior to bing intentionl. The sme asumption of intentionl
a
plogetics, controlled b the narrative voice, underlies the recent readin@ of the
Crin b Phillipp Ljeune in 1 pt autobiaphiu ( Pars, 1 976) and "L
peige casse," Patu 25 ( 1976) : 1-30.
EXCUSES ( CONFESSIONS ) 28
ever the content of the criminal act may have been, the excuse pre
sented in the conein was unable to stisf Russeau as a judge of
Jean-Jacques. This failure was already partly inscribed wthin the
excuse itself and it governs its further expansion and repetition.
Rousseau excuses himself from his gratuitous vciousness by
identitng his inner feeling as shm about himself rather than any
hostilit towards his vctim: ". . . the presence of so many people
was stronger than my repentance. I hardly feared punishment , my
only fear was shame; but I feared shame more than death, more
than the crime, more than anything in the world. I wished I could
have sunk and stifed myself in the center of the earth: unconquera
ble shame was stronger than anything else, shame alone caused my
impudence and the more gilt I became, the more the terror of
admitting my gilt made me fearless" (86) .
It is easy enough to describe how "shame" functions in a context
that seems to ofer a convncing answer to the question: what is
shame or, rather, what is one ashamed of Since the entire scene
stands under the aegis of thef, it has to do wth possession, and
desire must therefore be understood as functioning, at least at times,
as a desire to possess, in all the connotations of the term. Once it is
removed from its legitimate owner, the ribbon, being in itself devoid
of meaning and function, can circulate symbolically as a pure sig
nifer and become the articulating hinge in a chain of exchanges and
possessions . As the ribbon changes hands it traces a circuit leading to
the exposure of a hidden, censored desire. Russeau identifes the
desire as his desire for Marion: "it was my intention to give her the
ribbon" (86), i . e. , to "possess" her. At this point in the reading
sugested by Russeau, the proper meaning of the trope is clear
enough: the ribbon "stands for" Rousseau's desire for Marion or,
what amounts to the same thing, for Marion herself
Or, rather, it stands for the free circulation of the desire be
tween Rousseau and Marion, for the reciprocit which, as we kow
from juli, is for Rousseau the ver condition oflove; it stands for the
substitutabilit of Russeau for Marion and vce versa. Rousseau de
sires Marion as Marion desires Rousseau. But since, wthin the at
mosphere of intrige and suspicion that prevails in the household of
the Comtesse de Vercellis, the phantasy of this smmetrical reciproc
ity is experienced as an interdict , its fgure, the ribbon, has to be
stolen, and the agent of this transgression has to be susceptible of
being substituted: if Russeau has to be wlling to steal the ribbon,
ROUSSEAU
then Marion has t o be wlling t o substitute for Russeau i n perform
ing this act.
6
We have at least two levels of substitution (or displace
ment) takng place: the ribbon substituting for a desire which is itself
a desire for substitution. Both are governed b the same desire for
specular symmetr which gves to the smbolic object a detectable,
univocal proper meaning. The system work: "I accused Marion of
having done what 1 wanted to do and of having gven me the ribbon
because it was my intention to give it to her" (86) . The substitutions
have taken place wthout destroying the cohesion of the system,
refected in the balanced sytax of the sentence and now under
standable exactly as we comprehend the ribbn to signif desie.
Specular fgures of this knd are metaphors and it should be noted
that on this still elementar level of understanding, the introduction
of the fgural dimension in the text occus frst by ways of metaphor.
The allegor of this metaphor, revealed in the "confession" of
Russeau's desire for Marion, functions as an excuse if we are wlling
to take the desire at face value. If it is granted that Marion is desir
able, or Russeau ardent to such an extent, then the motivation for
the thef becomes understandable and easy to forgive. He did it all
out of love for her, and who would be a dour enough literalist to let a
little property stand in the way of young love? We would then b
willing to grant usseau that ''viciousness was never further from
me than at this cruel moment , and when 1 accused the hapless grl, it
is bizarre but it is true that my friendship for her was the cause of
my accusation" (86). Substitution is indeed bizarre (it is odd to take
a ribbon for a person) but since it reveals motives, causes, and
desires, the oddit is quickly reduced back to sense. The stor may b
a rebus or a riddle in which a ribbon is made to signif a desire, but
the riddle can be solved. The deliver of meaning is delayed but by no
means impossible.
This is not the only way, however, in which the text functions.
Desire conceived as possession allows for the all-important introduc
tion of fgural displacement: things are not merely what the seem to
be, a ribbon is not just a ribbon, to steal can be an act of love, an act
performed by Russeau can be said to be performed by Marion and,
in the process, it becomes more rather than less comprehensible, etc.
Yet the text does not stay confned wthin this pattern of desire. For
6. It is therefore consistent tht, when the scheme ends in disaster, Marion
would say: 'je ne voudrois pas etre a votre place" (85) .
EXCUSES (CONFBSSIONS )
one thing, to excuse the crime of thef dos not sufce to excuse the
worse crime of slander which, as both common sense and Ruseau
tell us, is much harder to accept . 7 Neither can the shame be ac
counted for by the hidden natue of the desire, as would be the case
in an oedipal situation. s The interdict does not weig ver heavly
and the reelation of Russeau's desire, in a public sitation that does
not allow for more intimate self-examination, hardly warrants such
an outbust of shame. More imporant than any of these referential
considerations, the text is not set up in such a way as to cout
sympathy in the name of Maron's erotic charm, a strateg which
Russeau uses wth some skll in many other instances includng the
frst part ofJuli. Aother form of desie than the desie of possssion
is operative in the latter part of the stor, which also bears the main
performative burden of the excuse and in which the crime is no
longer that of thef.
The obvious satisfaction in the tone ad the eloquence of the
passage quoted above, the easy fow of hyperboles (" . . . je la cra
gnois [la honte] plu que l mort, plus que I crime, plus que tout au
monde. J'aurois voulu m'enforcer, m'etoufer dans Ie centre
de la tere . . . " [86] ) , the obvous deligt with which the desie
to hide is being revealed, all point to another structure of desire
than mere possession and independent of the particula taget of
the desie. One is more ashamed of the exposure of the desie to
expose oneself than of the desire to possess; like Freud's dreams of
nadness, shame is primarily exhibitionistic. Wat Russeau rally
wanted is neither the ribbon nor Marion, but the public scene of
exposure which he actually gets. The fact that he made no attempt to
conceal the evidence confrms this. The more crime there is, the
more theft, lie, slander, and stubborn persistence i each of them,
the better. The more there is to expose, the more there is to be
ashamed of; the more resistance to exposure, the more satsfng the
scene, and, especially, the more stisfng and eloquent the belated
revelation, in the later narrative, of the inabilit to reveal. This desir
7. "To lie for one's own advantage i deceit , to lie for the bneft of another i
fraudulent, to lie in order to hn i slander; it i the worst knd of le" (ou
Rer, 1029).
8. The embarassing stor of Rusau's nection b Mme. de Vercelli, who i
dyng of a cancer of the breast , immediately precedes the stor of Mon, but
nothing in the text su
g
ests a concatenation that woud allow one to substtte
Marion for Mme. de Vercellis in a scene of nection.
ROUSSEAU
i s truly shameful, for it sugests that Marion was destroyed, not for
the sake of Russeau's savng face, nor for the ske of his desire for
her, but merely in order to provde him wi th a stage on which to
parade his disgace or, what amounts to the sme thing, to fnish
him wth a good ending for Bok II of his Consin. The structure
is self-perptuating, en abim, as is implied in its descrption as
exposure of the desire to expose, for each new stage in the unveiling
sugests a deeper shame, a geater impossibilit to reveal , and a
geater satisfaction in outwitting this impossibilit.
The structure of desire as exosure rather than as possession
explains why shame functions indeed, as it does in this text , as the
most efective excuse, much more efectively than greed, or lust , or
love. Promise is proleptic, but excuse is belated and always occurs
afer the crime; since the crime is exposure, the excuse consists in
recapitulating the exposure in the gise of concealment . The excue
is a rse which permits exposure in the name of hiding, not unlie
Being, in the later Heideger, reveals itself by hiding. Or, put difer
ently, shame used as excuse permits repression to function as revela
tion and thus to make pleasure and gilt interchangeable. Guilt is
forgiven because it allows for the pleasure of revealing its repression.
It fol lows that repression is in fact an excuse, one speech act among
others.
But the text ofers futher possibilities. The analysis of shame as
excuse makes evdent the strong lin between the performance o
excuses and the act of understanding. It has led to the problematics
of hiding and revealing, which are clealy problematics of cognition.
Excuse occurs wthin an epistemologcal twilight zone beteen
knowing and not-knowng; this is also why it has to be centered on
the crime of lying and why Rousseau can excuse himself for ever
thing provded he can b excused for lying. When this turns out not
to have been the case, when his claim to have lived for the se of
truth (vtm impe vr) is being contested from the outside, the
closu of excuse ("qu'il me soit permis de n'en rparler jamais")
becomes a delusion and the Fourh Rr has to be wtten.
The passge also stakes out the limits of how this understanding
of understanding then is to be uderstood. For the distinction be
tween desire as possession and desire as eposure, although it unde
niably is at work wthin the tet , does not structue its main move
ment . It could not be said, for instance, that the latter deconstructs
the former. Bth converge towars a unifed signifcation, and the
shame experenced at the desire to possess dovtails wth the deeper
EXCUSES ( CONFESSIONS ) 27
shame felt at self-exposure, just as the excuse for the one conspired
wi th the excuse for the other in mutual reinforcement. This implies
that the mode of cognition as hidingrevealing is fundamentally akin
to the moe of cognition as possession and that, at least up till this
pint , to know and to own are structured in the same way. Truth is a
prper of entities, and to lie is to steal , like Prometheus, this truth
away from its ower. In the devousness ofthe excuse pattern, the lie
is made legitimate, but this occurs wi thin a sstem of truth and
falsehoo that may be ambigous in its valorization but not in its
structure. It also implies that the terminolog of repression and
exposure encountered in the pasage on shame is entirely compatible
wth the sytem of symbolic substitutions (based on encoded sig
nifcations arbitrarily attributed to a free signifer, the ribbon) that
govern the passage on possessive desire ('1e l'accusai d'avoir fait ce
que je voulois faire . . . " [86]). The fgural rhetoric of the passage,
whose underlyng metaphor, encompssing both possession and ex
posure, is that of unveiling, combines wth a generalized pttern of
tropological substitution to reach a convncing meaning. Wat
seemed at frst like irrational behavior bordering on insanit has, by
the end of the passage, become comprehensible enough to b incor
porated within a general economy of human afectivt, in a theor
of desire, repression, and self-analyzing discourse in which excuse
and knowledge converge. Desire, now expanded far enough to in
clude the hidingrevealing movement of the unconscious as well as
possession, functions as the caue of the entire scene (" . . . it is
bizarre but true that my friendship for her was the caue of my
accusations" [86]), and once this desire has been made to appear in
all its complext, the action is understood and, consequently,
excused-for it was primarily its incongrit that was unforgvable.
Kowledge, moralit, possession, exposure, afectivit (shame as the
synthesis of pleasure and pain) , and the performative excuse are all
ultimately part of one system that is epistemologically as well as
ethically grounded and therefore available as meaning, in the mode
of understanding. Just as in a somewhat earlier passage of the Con
fesion the particular injustice of which Rousseau had been a vctim
becomes, by metaphorcal syecdoche, the paradigm for the univer
sal experience of injustice, 9 the episode ends up in a generalized
economy of rewards and punishments. The injur done to Maron is
9. Se the episoe of MIle Lmbrcier's broken comb in Bok I of the Confe
sin, especially p. 20.
28
ROUSSEAU
compensated for by the subsequent sufering inflicted on Rousseau
b nameless avengers acting in her stead. 10 The restoration of justice
naturally follows the disclosure of meaning. Wy then dos the ex
cuse fail and why does Rousseau have to return to an enigma that has
been so well resolved?
We have, of course, omitted from the reading the other sentence
in which the verb "excuser" is explicitly being used, again in a some
what unusual construction; the odity of "que je craignisse de m'ex
cuser" is repeated in the even more unusual locution: 'Je m'excusai
sur Ie premier objet qui s'ofrit" ("I excused myself upon the frst
thing that ofered itself' [86]), as one would say 'je me vengeai" or
'je m'acharnai sur Ie premier objet qui s'ofrit.,
,
1 1 The sentence is
inserted, it is true, wthin a context that may seem to confrm the
coherence of the causal chain: " . . . it is bizarre but it is true that my
friendship for her was the cause of my accusation. She was present to
my mind, 1 excused myself on the frst thing that ofered itself. 1
accused her of having done what 1 wanted to do and of havng gven
me the ribbon because it was my intention to give it to her . . . "
(86). Because Russeau desires Marion, she haunts his mind and her
name is pronounced almost unconsciously, as if it were a slip, a
segment of the discouse of the other. But the use of a vocabular of
contingenc ("Ie premier objet qui s'ofrit") within an argment of
causalit is arresting and disruptive, for the sentence is phrased in
such a way as to allow for a complete disjunction bteen Rous
seau's desires and interests and the selection of this particular name.
Marion just happened to be the frst thing that came to mind; any
other name, any other word, any other sound or noise could have
done just as well and Marion's entr into the discourse is a mere
efect of chance. She is a free signifer, metonymically related to the
10. "If this crime can b redeemed, as I hop it may, it must be b the many
misfortunes that have darkened the later part of my life, by forty years of upright
and honorable bhavior under difcut circumstances. Por Marion fnds so many
avengers in this world that , no matter how considerably I have ofended her, I have
little fear that I wll can this guil t wth me. This is ali I had to say on thi matter.
May I be allowed never to mention it agin" (87).
11. Te editor of the Pleiade Russeau, Marcel Rymond, comments on the
passage and quotes Rmon Fernandez (D 1 prsnnalit, p. 77): "He accuses her as
if he leaned on a piece of ftitr to avoid falling." Ryond spa of "an amot
dreamlie movement dictated by an unconsious which suddenly feels i tslf accused
and by which he transfers the 'misdeed' upon the other, on his nearb partner"
(1273) .
EXCUSES ( CONFESSIONS) 289
part she is made to play in the subsequent system of exchanges and
substitutions. She is, however, in an entirely diferent situation than
the other free sigifer, the ribbon, which also just happened to b
ready-to-hand, but which is not in any way itself the object of a
desire. Wereas, in the development that follows and that introuces
the entire chain leading from desire to shame to (dis)possession to
concealment to revelation to excuse and to distributive justice, Mar
ion can be the organizing principle because she is considered to be
the hidden center of an uge to reveal. Her bondage as target liberates
in turn the free play of her sybolical substitutes. Unlike the ribbon,
Marion is not herself divested of positive signifcation, since no reve
lation or no excuse would be possible if her presence within the chain
were not motivated as the target of the entire action. But if her
nominal presence is a mere coincidence, then we are entering an
entirely diferent system in which such terms as desire, shame, guilt,
exposure, and repression no longer have any place.
In the spirit of the text, one should resist all temptation to give
any signifcance whatever to the sound "Marion." For it is only if the
act that initiated the entire chain, the utterance of the sound "Mar
ion," is truly without any conceivable motive that the total arbitrari
ness of the action becomes the most efective, the most efcaciously
perormative excuse of all. The estrangement between subject and
utterance is then so radical that it escapes any mode of comprehen
sion. When everhing else fails, one can always plead insanit. "Mar
ion" is meaningless and powerless to generate by itself the chain of
causal substitutions and fges that strctures the surounding text,
which is a text of desire as well as a desire for text. It stands entirely
out of the system of truth, vrtue, and understanding (or of deceit,
evl, and eror) that gives meaning to the passage, and to the Cons
su as a whole. The sentence: ')e m'excusai sur Ie premier objet qui
s'ofrit" is therefore an anacoluthon, 1 2 a foreign element that disrupts
12. Classical rhetoric mentions anacoluthon espcially wth regrd to the
strcture of periodical sntences, when a shi f, stactical or other, occurs between
the frst part of the pero (protasis) and the second part (apodosis) . Heinrich
lusbrg i Hanuh d L. Ror (Munich, 19) , 1 :459, 92, gve
an example from Vergil : "quamquam anima meminis horrt luctuque rfugit ,
incipiam" (Ae 2, 12) . The following emple from. Rcine i frequently quoted:
"Vous voulez que ce Dieu vous comble de bienfaits / Et ne l'aimer jamais. "
Anacoluthon is not restricted to uninflected parts of speech but can involve nouns or
infected shi fers such as pronouns. It desigates any grammatical or sntactical
discontinuity in which a constrction interpts another bfore it is completed. A
29
ROUSSEAU
the meaning, the readabilit of the apologetic discourse, and reopens
what the excuse seemed to have closed of How are we to under
stand the implications of this sentence and what does it do to the
ver idea of understanding which we found to b so intimately
bound up wth and dependent upon the performativ function itself
The question takes us to the Fourh Rr and its implicit shif
from reported guilt to the gilt of reporting, since here the lie is no
longer connected wth some former misdeed but specifcally with the
act of writing the Coneion and, by extension, with all wting. Of
course, we always were in the realm of wting, in the narrative of
the Conein as well as in the Rer, but the thematization of this
fact is now explicit : what can b said about the interference of the
cognitive wth the performative function of excuses in the Fourh
Rr wll disseminate what existed as a localized disruption in the
Conein.
With the complicit of the casual, ambling, and free-associating
mode of the Rer, the text allows itself a puzzling lack of conclu
siveness. Cast in the tone of a pietistic self-examination, it sounds
severe and rigorous enough in its self-accusation to give weight to the
exoneration it pronounces upon its author-until Russeau takes it
all back in the penultimate paragraph which decrees him to be
"inexcusable" ( 1038) . There is also a strange unbalance between the
drift of the argument , which proceeds by fne distinctions and
ratiocinations, and the dri of the examples, which do not quite ft
their declared intent . The claim is made, for example, that, in the
Coneion, Russeau lef out several episodes because the showed
him in too favorable a light ; when some of these incidents are then
bing told in order to make the disfgured portrait more accurate,
they turn out to be curiously irrelevant . They do not show Rousseau
in all that favorable a light (since all he does is not to denounce
playing companions who harmed him by accident mid from whose
denunciation he would, at the time, have stood to gain ver little)
and the are, moreover, most unpleasant stories of physical assaul t,
b
loody mutilation, and crushed fngers, told in such a way that one
remembers the pain and the cruelt much better than the virtue the
are supposed to illustrate. All this adds to the somewhat uncanny
strikng instance of the strctural and epistemological implications of anacoluthon
occur in Proust in the desription of the lies used by Abrtine ("L prisonniere," A
l rhrh du tmp pru [Paris: Phiade, 1954] , 3: 153) . For Russau's ow
description of an anacoluthon-like si tuation, se note 16.
EXCUSES ( CONFESSIONS ) 291
obliqueness of a slightly delirious text which is far from mastering
the efects it pretends to produce.
The implications of the random lie in the Marion episoe ('Je
m'excusi sur Ie premier objet qui s'ofrit") are distributed, in the
Fourh Rr, over the entire te;t . The performative power of the lie
as excuse is more strongly marked here, and tied specifcally to the
absence of referential signifcation; it also carries , in this literar
context , a more familiar and reputable name since it is now called
fin: "To lie without intent and wthout harm to oneself or to
others is not to lie: it is not a lie but a fction" ( 1029). The notion of
fction is introduced in the sme way that the excuse of randomness
functions in the Conein. Within the airtight sstem of absolute
truth it produces the almost imperceptible crack of the purely
gatuitous, what Russeau calls "un fait oiseux, indiferent a tous
egrds et sans consequence pour personne . . . " ("a fact that is
totally useless , indiferent in all respects and inconsequential for
anyone" [1027]). There is some hesitation as to whether such "per
fectly sterile truths" are at all conceivable, or if we possess the neces
sar judgment to decide authoritatively whether certain statements
can be to that extent devoid of any signifcance. But although the text
vacillates on this point, it nevertheless functions predominantly as if
the matter had been settled positively: even if such truths are said to
be "rares et difciles," it is asserted that the "truth" of such "useless
facts" can be withheld without lying: "Truth deprived of any con
ceivable kind of usefulness can therefore not be something due [un
che du] , and consequently the one who keeps it silent or disguises
it does not lie" ( 1027). Moreover, "I have found there to be actual
instances in which truth can be withheld without injustice and dis
gised wi thout lyng" ( 1028). Some speech acts (although they might
better be called silence acts) therefore escape from the closed system
in which trth is property and lie thef: " . . . how could truths
entirely devoid of use, didactic or practical , be a commodity that is
due [un bin du) , since they are not even a commodit? And since
ownership is only based on use, there can be no property where there
can be no use" ("ou il n'y a point d' utilite possible il ne peut y avoir de
propriete" [1026] . Once this possibilit is granted, these fre-foating
"truths" or "facts," utterly devoid of value ("R ne peut etre dl de
ce qui n'est bon a rn" [ 1027]) are then susceptible of being "used"
as an excuse for the embellishments and exagerations that were
innocently added to the Coneion. The are mere "details oiseu"
ROUSSEAU
and t o call them lies would be, i n Russeau's words, "t o have a
conscience that is more delicate than mine" ( 1030) . The same para
gaph calls these weightless, air non-substance fctions: "whatever,
albeit contrar to truth, fails to concern justice in any way, is mere
fction, and I confess that someone who reproaches himself for a
pure fction as if it were a lie has a conscience that is more delicate
than mine" ( 1030) . Wat makes a fction a fction is not some polar
it of fact and representation. Fiction has nothing to do with repre
sentation but is the absence of any link between utterance and a
referent, regardless of whether this link be causal , encoded, or gov
erned by any other conceivable relationship that could lend itself to
systematization. In fction thus conceived the "necessar lin" of the
metaphor has been metonyized beyond the pint of catachresis,
and the fction becomes the disruption of the narrative's referential
illusion. This is precisely how the name of Marion came to be uttered
in the key sentence in the consion: 'e m'excusai sur Ie premier
objet qui s'ofrit ," a sentence in which any anthropomorphic conno
tation of seduction implied by the verb "s'ofrir" has to be resisted if
the efectiveness of the excuse is not to be undone and replaced by
the banali t of mere bad faith and suspicion. Rousseau was mang
whatever noise happened to come into his head; he was sayng no
thing at all , least of all someone's name. Because this is the case the
statement can function as excuse, just as fction functions as an
excuse for the disfgrations of the Consion.
It will be objected that fction in the Re and the denun
ciation of Marion are miles apart in that the former is without
consequence whereas the latter results in considerable damage to
others. Russeau himself stresses this: "whatever is contrar to trth
and hurts justice in any conceivable way is a lie" (1030) , and also
"the absence of a purpsefully harmful intent does not sufce to
make a lie innocent; one must also be assured that the error one
inficts upon one' s interlocutor can in no conceivable way harm him
or anyone else" ( 1029) . But the fction, in the coneion, becomes
harmful only because it is not understo for what it is, because the
fctional statement , as it generates the system of shame, desire, and
repression we described earlier, is at once caught and enmeshed in a
web of causes, signifcations, and substitutions. If the essential non
signifcation of the statement had been properly i nterpreted, if Rous
seau's accusers had realized that Marion's name was "Ie premier
objet qui s'ofrit ," the would have understood his lack of guilt as
well as Marion's innocence. And the excuse would have extended
EXCUSES (CONFESSIONS ) 29
from the slander back to the thef itself, which was equally unmoti
vated: he took the ribbon out of an unstated and anachic fact of
proximit, wthout awaeness of any law of owership. Not the
fction itself is to blame for the consequences but its falsely referen
tial reading. As a fction, the statement is innocuous and the error
harmless; it is the misgided reading of the eror as thef or slander,
the refusl to admit that fction is fction, the stubborn resistance to
the "fact ," obvious by itself, that langage is entirely free with regad
to referential meaning and can posit whateer its gam a alows it to
say, which leads to the transformation of random eror into injustice.
The radical irresponsibilit of fction is, in a way, so obvious, that it
seems hardly necessr to caution aginst its misreading. Yet its
assertion, within the stor of the Conesion, appears paradoxical
and far-fetched to the point of absurdit, so much so that Russeau's
own text , aginst its author's interests, prefers being suspected of lie
and slander rather than of innocently lackng sense. It seems to be
impossible to isolate the moment in which the fction stands free of
any signifcation; in the ver moment at which it is posited, as well as
in the context that it generates, it gets at once misinterpreted into a
determination which is, ipso fat, overdetermined. Yet wthout this
moment , never allowed to exist as such, no such thing as a text is
conceivable. We know this to be the case from empirical experience
as well: it is always poSSible to face up to any experience (to excuse
any gilt) , because the experience always exsts simutaneously as
fctional discourse and as empirical event and it is never possible to
decide which one ofthe to possibilities is the right one. The indeci
sion makes it possible to excuse the bleakest of crimes because, as a
fction, it escapes from the constraints of gilt and innocence. On the
other hand, it makes it equally possible to accuse fction-makng
which, in H6lderlin's words, is "the most innocent of all activities," of
being the most cruel . The knowledge of radical innocence also per
forms the harshest mutilations. Excuses not only accuse but they
can out the verdict implicit in their accusation.
This other aspect of radical excuse is also conveyed by the tet of
the Rer, though necessarily in a more oblique manner. In telling
another instance of a situation in which he lied out of shame-a less
interesting exmple than the ribbon, because there is nothing enig
matic about a lie which, in this case, is only a defense1 3-Russeau
13. In this case he is being provoked into lying by the half-teasing, half
malicious questions of a woman inquirng whether he ever had children.
ROUSSEAU
wt
es: "It is certain that neither my judgment, nor my wll dictated
my
reply, but that it was the automatic resut [l'et mhinl] of my
embarrassment" ( 1034) . The machinelik qualit of the text ofthe lie
is
more remakble still when, as in the Maon episode, the dispro
prtion btween the crime that is to b confessed and the crme
performed by the lie adds a delirious element to the sitation. By
saying that the excuse is not only a fction but also a machine one
adds to the connotation of referential detachment , of gatuitous im
provstion, that of the implacable repetition of a preordained pat
ter. Like Keist's marionettes, the machine is both "anti-gav," the
anamorphosis of a form detached from meaning and capable of
takng on any strcture whatever, yet entirely rthless in its inabilit
to moif its own strctural design for nonstrctural reasons. The
machine is like the grammar of the text when it is isolated from its
rhetoric, the merely formal element without which no text can be
generated. There can be no use of langage which is not, within a
certain perspective thus radically formal, i . e. mechanical, no matter
how deepl this aspect may be concealed by aesthetic, formalistic
delusions.
The machine not only generates, but also suppresses, and not
always in an innocent or balanced way. The economy of the Fourh
Rr is curiously inconsistent , although it is strongy thematized in
a text that has much to do wth additions and curtailments, with
"flling holes" ("rempIir les lacunes" [1035]) and creating them. The
parts of the text which are destined to be mere additions and exem
plifcations acquire autonomous power of signifcation to the point
where they can be said to reduce the main argment to impotence.
The addition of examples leads to the subversion of the cognitive
afrmation of innocence which the examples were supposed to illus
trate. At the end of the text , Russeau knows that he cannot be
excused, yet the text shelters itself from accusation by the perfor
mance of its radical fctionalit.
The literal censorship and cutailment of texts appears promi
nently in several places. A quotation from Tasso provdes a frst
example: Rousseau compares his ow resolve not to denounce his
playing companion to Sophronie's sacrifcial lie when, in order to
save the life of the Christians, she confessed to a crime (the thef ofa
rligous icon) that did not take place. The comparison borders on
the ludicrous, since Rousseau' s discretion is in no way equivalent to a
sacrifce. But the quotation which Rousseau now inserts into the text
EXCUSES ( CONFESSIONS ) 295
seres a diferent function. It is a passage which he had omitted,
without apparent reason, from the translation he made of the Sc
ond Canto of Tasso's epicl 4 at an earlier date. Any mention of Tasso,
in Russeau, always carries a high afective charge and generates
stories clustering around dubious translations , literar falsifcations,
textual distortions, fallacious prefaces as well as obsessions of iden
tifcation involving erotic fantasies and anxeties of insanit. 1 5 Limit
ing oneself, in this context, to the obvous, the insertion of the quota
tion must b an attempt to restore the integrit of a text wtten by
someone of whom Russeau himself had said "that one could not
suppress from his work a single stanz, from a stanza a single line,
and from a line a single word, without the entie poem collaps
ing . . . .
"1 6
But the restoration occurs as an entirely private and
secretive gesture, not unlike the citizen stealing "en secret" the word
"chacun" and thinkng of himself when he votes for all . 1 7 Such a
secretive reparation enforces the shameflness ofthe crime as well as
destroying any hope that it could be repaired. The mutilation seems
14. The translation is available in sveral of the early Rousseau edi tions, for
example in Ouvr complte d J. J. Rus au (Au deux pnts: chez Snson et
Compagnie, 1792) , 4: 215-47. It is printed in bilingual version and even the early
editors had obsered and indicated the absence of the passge which was later to b
quoted in the Fourth Rer (ibid. , 229) .
15. On Rousseau and Tasso, one fnds general obserations, not vel infon
tive in this context , in several articles, mostly by Italian authors, mentioned by
Berard Guyon in his notes to the Pleiade edition of the Nouvll Heli (2: 1339) .
16. The statement is not a quotation from Russeau but is reported by Coran
cez in DJ. J. Rous au (Extrait dUJournal de Paris, # 251 , 256, 259, 260, 261 , An 6,
42-43) . The sequel of the statement , in which Russeau descrbes t he one exception
to the organic integrit of Tasso's work, is equally interesting for our purposes and
could b read as Rousseau's description of an anacoluthon: " . . . sans que Ie pme
entier ne s'ecroule, tant (Ie Tasse) etait precis et ne mettait rien que de necessaire. Eh
bien, otez la strophe entiere dont je vous parle; rien n'en soufre, l'ouvrage reste
parfait. Elle n'a rapport ni a ce qui precede, ni a ce qui sui t; c'est une pii-e absolu
ment inutile. II est a presumer que Ie Tasse l'a fite involontaiement et sns l
comprendre lui-meme; mais elle et claire. " Corancez could not remembr the
stanza Rousseau quoted, but it has been tentatively identifed as stanza 77 of Canto
XII of Jelm Dlivrd. S L. Proal . L pschlo d J. J. Rus au (Pars: F.
Alcan, 1923) , p. 327 and Ouve complte, 1 : 1386-87, which Rousseau chos to
read as the prefguration of his ow prsecutions. Corancez tells the stor as an
instance of Russeau's growng paranoia and, in the sme article, he reports Rous
seau's death as suicide. His article is wi tten in defens, however, of Russau's
memol.
17. Sil Contat (3:306) .
ROUSSEAU
t o be ic
ab
le and the prothesis only serves t o mark this fact more
stron
g
y
.
The accusation
that hangs over the entire Fourh Rer
and again
s
t
which the excuse tries to defend itself seems to have to
do wt
h a thre
a
t oftetual mutilation, itselflinked to the organic and
totalizing synechdocal language by means of which Russeau refers
to t
he uni t of Ta
sso's work.
The omission and surreptitious replacement of the Sophronie
passage is at most a syptom, all the more so since "Tasso," in
Rousseau, implies a threat as well as a vctim, a weapon as well as a
wound. The mutilation is not just the excision of one specifc piece of
text . Its wder signifcance becomes more evident in another literar
allusion in the Fourh Pomna, the reference to Montesquieu's
conventionally deceptive preface to L Temple d Gni. By pretend
ing that his work is the translation ofa Greek manuscript , the author
shelters himself from the possible accusation of frivolity or licen
tiousness, kowng that the reader who is enlightened enough not
to hold his levty aginst him wll also be sufciently informed about
literar convention not to be taken in by the phony preface. Russeau
treats Montesquieu's hoax without undue severity ("Could it have
occurred to anyone to incriminate the author for this lie and to call
him an impostor? [ 1030]), yet behind this apparent tolerance stands
a much less reassuring question. As we know from the "Preface
dialoguee" to the Nouvelle Heloie, the preface is the place in the text
where the question of textual master and authority is being decided
and where, in the instance ofuli, it is also found to be undecidable.
With this threatening loss of control the possibility arises of the
entirely gatuitous and irresponsible text , not just (as was apparently
the case for Montesquieu or for naIve readers ofuli) as an inten
tional denial of paternity for the sae of self-protection, but as the
radical annihilation of the metaphor ofselfood and of the wll . This
more than warrants the anxiety wth which Rousseau acknowledges
the lethal qualit of all writing. Writing always includes the moment
of dispossession in favor of the arbitrar power play of the signifer
and from the point of vew of the subject, this can only be experi
enced as a dismemberment , a beheading or a castration. Behind
Montesquieu's harmless lie, denyng authorship of L Temple d
Gni by the manipulation of the preface that "heads" the text ,
stands the much more dangerous ambivalence of the "beheaded"
author.
I I
18. The same anxiet i s a
pp
arent in another rfernce t o
p
rfaces i n Russeau,
EXCUSES ( CONFESSIONS ) 297
But precisely because, in all these instances, the metaphor for
the text is still the metaphor of text as body ( from which a more or
less vtal part , including the head, is being severed) , the threat re
mains sheltered behind its metaphoricity. The possible loss of au
thorship is not wthout consequences, liberating as well as threaten
ing, for the empirical author, yet the mutilation of the text cannot be
taken seriously: the clear meaning of the fge also prevents it from
carrng out what this meaning implies. The undecidability of au
thorship is a cogition of considerable epistemological importance
but , as a cognition, it remains ensconced within the fgural delusion
that separates kowng from doing. Only when Rousseau no longer
confronts Tasso's or Montesquieu's but his ow text , the Conin,
does the metaphor of text as body make way for the more directly
threatening alternative of the text as machine.
Unlike the other two texts, where the distortion had been a
suppression, the Consin is at frst guilty of disfguring by excess,
by the addition of superfuous, fctional embllishments, "I have
nevr said less, but I have sometimes said more . . . " ( 1035) , but a
few lines later it turns out that this was not the case either, since
Russeau admits having omitted some of his recollections from the
narrative merely because they showed him in too favorable a light .
There is less contradiction between the to statements when it turns
out that what he omitted are precisely stories that narrate mutila
tions or, in the metaphor of the text as body, suppressions. Both
stories have to do wth mutilation and beheading: he nearly loses a
hand in the fst and comes close to havng his brains kocked out in
the other. Thus to omit suppressions is, in a sense (albeit by syllep
sis) , to preserve an integrit, "ne jamais dire moins. " If the stories
that have been omitted threaten the integrit of the text , then it
interestingly enough also in connection with Tasso. To deny authorship in a preface
in the name of trth (as Russeau did in the cae ofuli) do not only mean that
one's authorship of all texts can be put in question but as that all texts can b
attributed to one. This i precisely what happns to Russeau when a malevolent (or
commercially enterprising) editor, in what reads like a transparent parody of the
"Pae dilgee," attributes to him a poor translation of Tass'sJeram Deliv
er, (see Oeuve complt, 1 : 1740 for the text of the editor's prae, and also
Ouve complt, 1 : 138) . Russau mentions the' incident with some degee of
paranoid aniety in a letter to Mme. de Lssert of August 23, 1774, and among many
other instances of fals textual attribution, in the Dil (90) . The chain that
leads from Tass to translation, to prefaces, to authorship, to bheading, and to
insnity is ready to surface in any context of anxiety about trth and falshod.
ROUSSEAU
woud be even easier to excuse him for not havng included them
tha to excuse
pim for the superuous ornaments he added to the
recolection of his happier memories.
But in what way are these narratives threatening? A instances
of Russeau's generosit the are, as we already pointed out, more
inept than convincing. They seem to exst primarily for the sake of
the mutilations they describe. But these actual, bodily mutilations
seem, in their turn, to be there more for the sae of allowng the
eocation of the machine that causes them than for their ow shock
value; Russeau lingers complacently over the description of the
machine that seduces him into dangerously close contact: "I looked
at the metal rolls, my eyes were attracted by their polish. 1 was
tempted to touch them with my fngers and I moved them with
pleasure over the polished surface of the clinder . . . " (1036) . In the
general economy of the Rr, the machine displaces all other sig
nifcations and becomes the raison d'etre of the text. Its power of
sugestion reaches far beyond its illustrative puspose, especially if
one bears in mind the prevous characterization of unmotivated,
fctional language as "machinal." The underlying strctural patterns
of addition and suppression as well as the fgral system of the text al
converge towards it. Barely concealed by its peripheral function, the
text here stages the textual machine of its ow constitution an
performance, its ow textual allegor. The threatening element in
these incidents then becomes more apparent. The text as boy, with
all its implications of substitutive trops ultimately always retrace
able to metaphor, is displaced b the text as machine and, in the
process, it sufers the loss of the illusion of meaning. The deconstruc
tion of the fgural dimension is a process that taes place indepen
dently of any desire; as such it is not unconscious but mechanical,
systematic in its performance but arbitrar in its principle, like a
gammar. This threatens the autobiographical subject not as the loss
of something that once was present and that it once possessed, but as
a radical estrangement between the meaning and the prformance
of any text.
In order to come into being as text, the referential function had
to be radically suspended. Without the scandal of random denuncia
tion of Marion, wthout the "faits oiseux" of the Confsion
,
there
could not have been a text; there would have been nothing to excuse
since everthing could have been explained away by the cognitive
logic of understanding. The cognition would have been the excuse,
EXCUSES ( CONFESSIONS )
299
and this convergence is precisely what is no longer conceivable as
soon as the metaphorical integit of the text is put in question, as
soon as the text is said not to be a fgral body but a machine. Far
frm seeing language as an instrument in the servce of a psychic
energ, the pssibilit now arises that the entire construction of
drives, substitutions, repressions, and representations is the aber
rant, metaphorical correlative of the absolute randomness of lan
gage, prior to any fguration or meaning. It is no longer certain that
language, as excuse, exists because ofa prior guilt but just as possible
that since language, as a machine, prforms anyway, we have to
produce guilt (and all its train of psychic conseqences) in order to
make the excuse meaningful. Excuses generate the ver guilt they
exonerate, though always in excess or by default. At the end of the
Rr there is a lot more guilt around than we had at the start :
Russeau's indulgence in what he calls, in another bodily metaphor,
" plaiir d'ecrire" ( 1038) , leaves him gltier than ever, but we now
have also the two companions of his youth, Pleince and Fazy, guilt
of assault , brtalit or, at the ver best , of carelessness. 1 9 Additional
gilt means additional excuse: Fazy and Pleince now both have to
apologize and may, for all we kow, have written moving texts about
the dreadful things the did to Jean-acques who, in his turn, now
has to apologize for having possibly accused them arbitrarily, as he
accused Marion, simply because their names may have happened to
ocur to him for the least complling ofreasons.2o No excuse can ever
hope to catch up with such a proliferation of gil t . On the other
hand, any guilt , including the guilt pleasure of writing the Fourh
Rr, can always b dismissed as the gratuitous product of a
textual gammar or a radical fction: there can never be enough guilt
around to match the text-machine's infnite power to excuse. Since
gilt , in this description, is a cognitive and excuse a performative
function of language, we are restating the disjunction of the perfor-
19. The description of the way in which Fazy iIjured Russau is ambiguous,
since the narative is phrased in such a way that he can b suspcted of having done
i t wth delibration: " . . . Ie june Fazy s'etant mis dans la roue lui donna un
demiquart de tour si adroitement qu'il n'y prit que Ie but de mes deux plus longs
doigts; mais e'en fut assz pou qu'ils fussnt &rases . . . " ( 106) .
20. For example, the fact that their names may have come to mind becaus of
their phonic resemblance to the place names where t he incidents are sid to have
taken place: the one involving Fazy ocurs at Paquis, the one involving Pleince at
Plain-Palais.
3
ROUSSEAU
mative from the cogitive: any speech act produces an excess of
cognition, but it can never hope to know the process of its own
production (the only thing worth kowing. Just as the text can never
stop apologizing for the suppression of guilt that it prforms, there
is never enough kowledge available to account for the delusion of
knowing.
The main point of the reading has been to show that the result
ing prdicament is linguistic rather than ontologcal or hermeneutic.
As was clear from the Marion episode in the Confsion, the decon
strction of tropological patterns of substitution (binar or ternar)
can be included wthin discourses that leave the assumption of intel
ligbility not only unquestioned but that reinforce this assumption by
making the mastering of the tropological displacement the ver bur
den of understanding. This project engenders its ow narrative
which can be called an allegor of fge. This narrative begins to
vacillate only when it appears that these (negative) cogitions fail to
mae the performative function of the discourse predictable and
that, consequently, the linguistic model cannot be reduced to a mere
system of tropes. Performative rhetoric and cogitive rhetoric, the
rhetoric of tropes, fail to converge. The chain of substitutions
functions next to another, diferently structure system that exists
independently of referential determination, in a system that is both
entirely arbitrar and entirely repeatable, like a grammar. The inter
section of the two systems can b located in a text as the disrption
of the fgural chain which we identifed, in the passage from the
Confeion, as anacoluthon; in the language of representational
rhetoric, one could also call it parabasis,2 J a sudden revelation of the
discontinuity between two rhetorical codes. This isolated textual
event, as the reading of the Fourh Re shows, is disseminated
throughout the entire text and the anacoluthon is extended over all
the points of the fgral line or allegor; in a slight extension of
21 . The similarit bten ancoluthon and parabsis stems from the fact
that bth figures interrupt the expctations of a given grammatical or rhetorical
movement . As digssion, aide, "inttn d'auteur," or "au d Rll fall,"
parabsis dearly involves the interuption ofa discour. The quotation from Fried
rich Shlegl appar among the formerly unavailable note contemporar wth the
Lycum and Atheneum Fragmenten. Fierich Shlegel, Ktih Frh-Shlel
Au
g
b, ed. Ernst Bhler ( Munich, 19), 18: 85, 6. The us of the term parabasis
( or parekbasis) by Shlegel ehos the use of the devce espcially in the plys of
Tieck.
EXCUSES ( CONFESSIONS ) 301
Friedrich Schlegel's formulation, it becomes the permanent
parabasis of an allegor (of fge) , that is to say, irony. Irony is no
longer a trope but the undoing of the deconstructive allegor of all
tropologcal cognitions, the sstematic undoing, in other words, of
understanding. As such, far from closing of the trpologcal system,
irony enforces the reptition of i ts abrration.

Index
Althusser, Louis, 138, 224
Aristotl e, 120, 146, 232
Auerbach, Erich, 81
Austin,john L. , 7, 13, 281
Bachelard, Gaston, 1 7
Barthes, Roland, 6, 1 14
Baudelaire, Charles, 22, 49, 168, 189
Beardsley, Monroe, 10
Benjami n, Walter, 81
Blake, Wil liam, 18, 221
Burgel i n, Pierre, 223, 225
Burke, Kenneth, 8
Celan, Paul , 48
Cl arke, Samuel , 228
Condillac, Eti enne Bonnot de, 147, 149,
250
Curtius, Ernst Rbrt , 104
Dante Alighieri , 17, 83
Deleuze, Gilles, 72, 77, 83, 8
Derrida, jacques, 9, 82, 8, 139, 150,
180
Descartes, Rene, 95, 1 52, 227
Diderot , Denis, 1 67, 226, 230, 231 , 237,
250, 260
Ducrot , Oswald, 6
Eliot , T. S. , 20
Emerson, Rlph Waldo, 103
Fenelon, Fran<ois de Salignac de la
Mothe, 235
Fichte, johann Gottlieb, 189
Flaubert , Gustave, 226
Foucault , Michel , 5, 82, 146-48
Frege, Got tlob, 13
Freud, Sigmund, 82, 174, 175, 187, 244,
285
Genette, Grard, 6-8, 68, 72, 77
Gothe, johann Wolfgang von, 12, 73,
1 03, 1 76, 181 , 182, 215
Got h, joachi m, 104
Greimas, Algirdasjulien, 6
Grosrichard, Alai n, 146-49, 163
Guyon, Brnard, 189
Hamann, johann Gorg, 176
Hazl i tt , Wi lliam, 163, 194
Hegel , G. W. F. , 80, 82, 8
Heideger, Marti n, 24, 82, 88, 175, 187,
286
Helvetius, Claude-Adrien, 226, 250
Herder, johann Gottfried, 181
Hobbes, Thomas, 142, 250, 255
Holbach, Paul Henri Dietrich, baron d' ,
226
Holderlin, Friedrich, 17, 49, 103, 1 1 7,
1 97, 251 , 293
jakobson, Roman, 5, 6, 146
Kant , I mmanuel , 86, 179, 187, 189, 208,
239
Kleist , Heinrich von, 167, 185, 245, 294
Kofman, Sarah, 89, 10
Lados, Pierre Choderlos de, 193
Lacoue-Labarthe, Philippe, 83, 89, 96,
100, 104
La Mettrie, julien Ofroy de, 226, 235,
237
Leiris, Michel , 278
Locke, john, 147, 234, 235
Machiavell i , Niccol o, 1 1 3, 275
Malebranche, Nicolas de, 239
Mallarme, Stephane, 87, 200, 237, 268
Marx, Krl, 6, 82, 158
303
304
Milton, john, 1 7
Montaige, Michel de, 1 03, 227
M
onte
s
q
u
i eu, Charles de Secondat de,
82, 296
, 297
M6rchen, Hermann, 25, 26, 54
Nietzsche, Friedrich, 8, 9, 15, 1 7, 79-
1 31 , 1 54, 226, 244, 252
-Th Birh ofTragedy, 79-102, 1 10,
1 16-18, 131
-Th Gay &ience, 86, 116
-A Genealo ofMorls, 83, 1 1 6, 126,
127, 131
-Human al l too Human, 82, 106
--n Truth and Lie in an Extr-Morl
Sn, 1 1 0, 1 1 1 , 1 1 3, 1 1 5-18;
"Philosophenbuch," 83, 94, 1 06, 1 10,
1 1 3, 1 16, 1 1 7
-ThI Spak ZarthItr, 1 1 6
-Unzitgemise Betrahtungen, 1 1 7
-Th Will to Power, 106, 109, 1 1 1 , 1 13,
1 16
Ohmann, Richard, 8
Pautrat, Bernard, 104
Peirce, Charles Sanders, 8, 9
Petrarca, Francesco, 195, 20
Plato, 88, 1 03, 109, 130, 226
Poulet , Georges , 57, 1 90
Proust , Marcel , 4, 5, 13-19, 57-78, 189,
226, 262
Quesnay, Francois, 250
Rymond, Marcel , 1 90
Richter , jean Paul , 106
Ricoeur, Paul , 1 74, 1 75, 187
Rilke, Ri ner Maria, 20-56
-Das Buh der BUer: "Am Rnd dr
Nacht," 33-38; "Ei ngang," 39-40
-Das Stunden-Buh: Yom monchichen
Lben, "Ich liebe dich, du snfestes
Gesetz, " 28-33
-Die Snette an O"pheIS, 23, 49; "Sieh
den Himmel . Hei sst kei n Sternbild
' Reiter' ?" 51 -54
I NDEX
-Duineser Elegen, 23, 49, 50, 52
-Ltzte Gedichte und Frgn-
tarLchs: "Gong," 55; "II faut fermer
IIs yeux et renoncer a la bouche ,"
55-56
-Neu Gedichte: "Archaischer Torso
Apollos, " 4; "Der Ball , " 43-44, 54;
"Die Fensterrose," 45; "L'ange du
meridien, " 43; "Orpheus. Eurdike.
Hermes," 46-47; "Quai du Rosaire,"
40-43
Roenbach, Georges, 41
Rohde , Ern, 88
Rousseau, jeanjac
q
ues, 75, 82, 83, 8,
90, 103, 1 13, 122, 135-301
-onfsions, 147, 167, 248, 251 ,
278-301
-L Contrat social (Scial Contrat),
(Two versions: a. Geneva Manus
cript , b. Final Version) , 137, 1 57, 1 59,
1 62, 183, 209, 220, 223, 224, 225,
246-77, 278
-ialoges, 163
-DLcours sur l 'origne de l 'inigalite
pari les hommes, "Second Dis
course," 8, 93, 135-59, 1 60-68, 1 71 ,
1 72, 188, 197, 198, 200, 201 , 205,
206, 212, 216, 230, 231 , 233, 252,
266, 274, 275
-Du bonhur publi, 225, 257, 258,
260, 272
-mile, 152, 224, 226, 248
-Essai sur l 'origne des langs, 84,
140, 1 4, 147, 149, 1 53, 1 56, 1 62,
164, 167, 169, 197, 200, 201 , 227,
233, 239
-Julie au la Nouvelle HeloLe, 135, 1 52,
157, 15, 1 70, 188-220, 224, 246,
247, 248, 253, 257, 283, 296; Preface
toJuli, 1 97, 1 98, 202-10, 21 7, 237,
296
-Lettre a Frnquires, 225
-Lttres morls, 243, 244
-arcLse, 164-68, 1 70-72, 1 75, 1 76,
1 79, 183-85, 1 87, 210, 213
-Profsion d foi du vimi,.e Savoyard,
220, 221-45, 247, 250, 253, 257,
259-60, 275
I NDEX
-Pgalion. 160-87, 188, 210, 213,
214
-Rries d' un prmeneur solitair,
"Fourth Promenade," 136, 162, 187,
251 , 278, 279, 282, 286, 290, 292,
293, 294, 296, 298, 299, 300
Ruski n, john, 75
Si nte Beuve, Charles A. , 6
Saussure, Ferdinand de, 5, 8, 88
Schiller, Friedrich von, 137, 176, 190,
208, 232
Schlegel, Friedrich, 106, 1 16, 130, 301
Schopenhauer, Arthur, 86, 96, 100, 1 1 7
StaeI , Germaine de, 163
Starobi nski , jean, 141 -42, 1 43-44, 163,
1 67, 171 -72, 191 , 193
Stendhal , 1 14, 19
Sterne, Laurence, 1 03
Tasso, Torquat o, 195, 209, 294-97
Todorov, Tzvetan, 6, 8
Trakl , Georg, 20, 48
Tronchin , jean-Robert , 223
Valer, Paul , 5, 168
Voltaire, Francois M. A. , 221
305
Wagner, Rchard, 84-7, 90, 92, 95-97
Wi ttgenstein, Ludwig, 17, 1 1 1
Wordswort h, William, 192
Yeats, William Butler, 1 1-12, 20

S-ar putea să vă placă și